The Human GLP-1 Analogs Liraglutide and Semaglutide: Absence of Histopathological Effects on the Pancreas in Nonhuman Primates

2486 Diabetes Volume 63, July 2014

Carsten F. Gotfredsen, Anne-Marie Mølck, Inger Thorup, Niels C. Berg Nyborg, Zaki Salanti, Lotte Bjerre Knudsen, and Marianne O. Larsen

The Human GLP-1 Analogs Liraglutide and Semaglutide: Absence of Histopathological Effects on the Pancreas in Nonhuman Primates

Diabetes 2014;63:2486–2497 | DOI: 10.2337/db13-1087

Increased pancreas mass and glucagon-positive ade- and glucagon-like peptide-1 (GLP-1) receptor agonists nomas have been suggested to be a risk associated (GLP-1RAs). This concern is partly based on the well- with sitagliptin or exenatide therapy in humans. Novo described effect of GLP-1 to induce growth of pancre- Nordisk has conducted extensive toxicology studies, atic b-cells (1). Both drug classes increase effective including data on pancreas weight and histology, in GLP-1 levels, but to a different degree, and have different Cynomolgus monkeys dosed with two different human modes of action; thus, it is important to differentiate glucagon-like peptide-1 (GLP-1) receptor agonists. In a between them, particularly when considering mechanistic 52-week study with liraglutide, a dose-related increase hypotheses for potential safety concerns or signals. GLP- in absolute pancreas weight was observed in female 1RAs mediate their effects directly through the GLP-1R monkeys only. Such dose-related increase was not (2,3). Although increased levels of GLP-1 and glucose- found in studies of 4, 13, or 87 weeks’ duration. No dependent insulinotropic peptide (GIP) are considered treatment-related histopathological abnormalities were important parts of the mechanism of action of DPP-4is observed in any of the studies. Quantitative histology of (4), DPP-4 is known to degrade many other hormones (5).

PHARMACOLOGY AND THERAPEUTICS the pancreas from the 52-week study showed an in- DPP-4is have been shown to increase GLP-1, GIP, and crease in the exocrine cell mass in liraglutide-dosed peptide YY (6,7). This complicates the understanding of animals, with normal composition of endocrine and exo- crine cellular compartments. Proliferation rate of the both desired and potentially undesired effects of this class exocrine tissue was low and comparable between of compounds. Within GLP-1RAs, differences exist in groups. Endocrine cell mass and proliferation rates were safety-related parameters. One subgroup is the exendin- – unaltered by liraglutide treatment. Semaglutide showed 4 based drugs with exenatide and lixisenatide, which are no increase in pancreas weight and no treatment- structurally distinct from human GLP-1. Owing to the low related histopathological findings in the pancreas after amino acid homology to native human GLP-1, these med- 13 or 52 weeks’ dosing. Overall, results in 138 nonhuman ications are associated with an increased number of im- primates showed no histopathological changes in the mune reactions that are, however, all of a relatively mild pancreas associated with liraglutide or semaglutide, form, that is, mostly antibody development, injection two structurally different GLP-1 receptor agonists. site nodules, and loss of efficacy (8,9). The other sub- group is based on human GLP-1 and contains liraglutide, Pancreas safety has become a subject of much debate con- taspoglutide, and larger covalently conjugated molecules cerning dipeptidylpeptidase-4 (DPP-4) inhibitors (DPP-4is) such as albiglutide and dulaglutide. Clinical development

Diabetes Research Unit, Novo Nordisk, Måløv, Denmark © 2014 by the American Diabetes Association. See http://creativecommons.org Corresponding author: Lotte Bjerre Knudsen, [email protected]. /licenses/by-nc-nd/3.0/ for details. Received 11 July 2013 and accepted 18 February 2014. See accompanying article, p. 2219. This article contains Supplementary Data online at http://diabetes .diabetesjournals.org/lookup/suppl/doi:10.2337/db13-1087/-/DC1. diabetes.diabetesjournals.org Gotfredsen and Associates 2487 of taspoglutide was stopped due to severe immune-related RESEARCH DESIGN AND METHODS side effects, with cases of anaphylactic shock, possibly The research design and methods for liraglutide stud- fi caused by the formulation (8,10). No such ndings have ies in Cynomolgus monkeys have been described pre- – been reported with other GLP-1 based analogs. viously (13). All animals were examined daily for Semaglutide is a once-weekly GLP-1 analog that is in clinical signs in the in-life phase. Studies with sema- phase 3 clinical development (11). Where liraglutide is glutide were generally performed similarly, and both acylated with a palmitic acid and has an extra amino followed international guidelines provided by Interna- acid as a spacer between the palmitic acid and the Lys26, tional Conference of Harmonization. Dose levels, du- where the fatty acid is attached, semaglutide is acylated ration, and group sizes for all studies are described in with a stearic diacid at Lys26 but has a much larger syn- Fig. 1. Compounds were administered as subcutaneous thetic spacer and is, furthermore, modified for DPP-4 sta- injections. bility in position 8, where the amino acid a-aminobutyric acid has been introduced. Body Weight and Pancreas Weight In support of drug development and regulatory ap- Terminal body weight was obtained in sedated animals proval for treatment of chronic diseases, repeat dose immediately before they were killed. The entire pancreas toxicity and carcinogenicity studies are conducted. These of each animal was examined macroscopically for any ab- studies are performed at different dosing durations to normalities, excised, cleaned of fat and connective tissue, support the different phases of clinical development and and weighed. with doses aiming to obtain exposure several multiples higher than the clinically relevant exposure with the aim Tissue Preparation to identify potential drug-related organ toxicity and A transverse section from themidpartofthepancreas carcinogenicity. Repeat dose toxicity studies are typically from all animals was sampled, and this section and the fi conducted in a rodent and a nonrodent species. For rest of the pancreata were xed in 10% neutral buffered liraglutide and semaglutide, Cynomolgus monkeys were formalin for at least 48 h. The sections were dehydrated fi chosen as the nonrodent species. Repeat dose toxicity and paraf n-embedded according to standard histopath- studies in nonhuman primates are designed to screen ological procedures. According to international standard for potential hazards and are not designed or statistically practices, one section per animal was cut at a nominal thickness of 4–5 mm and stained with hematoxylin and powered for identification of differences in the incidence eosin (17). The slides were read unblinded because this is or severity of individual organ changes. The number of recommended by toxicopathology experts as a way to nonhuman primates per group is limited to three to five increase the chance of separating subtle changes from for ethical reasons (12). Because of the statistical limita- normal background changes (18). Further details of tions, standardization of the examinations is critical: If the methodology have been previously described (13). pathological findings are identified at a frequency or se- The pancreas specimens from animals killed at termina- verity exceeding those in the in-study control group, they tion of dosing in the 52-week liraglutide study were are often compared with historical control data to assist a fi fi evaluated by quantitative histology (all groups for -cells interpreting the signi cance of the nding. This principle and Ki-67, and only vehicle and high-dose groups for b-, fi also applies to organ weights. Histopathological ndings d-, pancreatic polypeptide [PP], ductal, and acinar cells). in the pancreas from the repeat-dose studies in rodents The cranial and caudal remnants were sectioned longi- and nonhuman primates and from carcinogenicity studies tudinally, cut into ;40 pieces, and distributed to four in rodents with liraglutide have been published previously capsules, with one-fourth in each, according to the (13). Liraglutide was not found to have a causal relation- smooth fractionator principle (19,20). Processing and fi ship to any histopathological ndings. staining for detection of b-, a-, d-, and PP cells was Some studies in rats and mice have shown an increased done as previously described (21). The reactivity of the pancreas weight induced by DPP-4is or GLP-1RAs (14,15), primary antibodies to insulin, glucagon, somatostatin, and a recent ex vivo study with human pancreata sug- and PP in monkey pancreas had been tested indi- gested an increase in glucagonomas as well as increased vidually to validate the method. Costaining for duct pancreas weight (16). Here, pancreas weight in Cynomol- plus acinar cells using mouse anti–CK-7 (Dako, Glostrup, gus monkeys is reported for liraglutide in toxicology stud- Denmark) and rabbit anti–a-amylase (Calbiochem, Darm- ies with 4, 13, 52, and 87 weeks’ dosing, and for stadt, Germany) plus Ki-67 using polyclonal rabbit semaglutide in toxicology studies with 13 and 52 weeks’ anti–Ki-67 (Novus Europe, Cambridge, U.K.) followed dosing, as well as a full histopathological evaluation of thesameprinciplesasabove.Stainedslideswere these same studies, except the 87 weeks’ study, which scanned in a Hamamatsu NanoZoomer 2.0-HT (Hamamatsu, has been reported previously (13). For liraglutide, a full Hamamatsu City, Japan). Images were subsequently ana- quantitative histological assessment of the endocrine as lyzed automatically in the Visiopharm Integrator System well as the exocrine pancreas was also performed in the (Visiopharm A/S, Hørsholm, Denmark) image analysis 52 weeks’ study. program. 2488 GLP-1 Receptor Agonists and the Pancreas Diabetes Volume 63, July 2014

Figure 1—Absolute pancreas weight (g) in male (M) and female (F) control and liraglutide (subcutaneous once daily) (A–D) or semaglutide (subcutaneous twice weekly) (E and F) dosed nonhuman primates (horizontal line indicates group mean in each data set). Liraglutide dosing for 4 weeks (A), 13 weeks (B), 52 weeks (C), and 87 weeks (D). Semaglutide dosing for 13 weeks (E) and 52 weeks (F). Con M and Con F are historical control animals in panel A–C (from other studies of the same duration, run in the same facility). 0M and 0F are vehicle-dosed animals on all graphs. A–C: 1M and 1F were dosed with 0.05 mg/kg/day, 2M and 2F with 0.5 mg/kg/day, and 3M and 3F with 5.0 mg/kg/day liraglutide. D: 1M and 1F were dosed with 0.25 mg/kg/day and 2M and 2F with 5.0 mg/kg/day liraglutide. §Animal killed at 63 weeks on humane grounds. In the 13-week semaglutide study, the high dose was reduced from 0.98 to 0.47 mg/kg after 6 weeks due to severe dehydration in 2 female animals indicating that 0.98 mg/kg was not a tolerated dose. These two animals were killed after 36 days and not included in the analysis of pancreas weight, leading to a group size of two. E: 1M and 1F were dosed with 0.004 mg/kg, 2M and 2F with 0.086 mg/kg, and 3M and 3F with 0.47 mg/kg semaglutide twice weekly. F: 1M and 1F were dosed with 0.01 mg/kg, 2M and 2F with 0.06 mg/kg, and 3M and 3F with 0.36 mg/kg semaglutide twice weekly. A:Males:P = 0.036 by ANOVA across M0, M1, M2, and M3; P = 0.054 vs. 0M for 1M by posttest; P = 0.18 by ANOVA across all groups, including Con M. Females: P = 0.38 by ANOVA across F0, F1, F2, and F3; P = 0.10 by ANOVA across all groups. B:Males:P = 0.73 by Kruskal-Wallis test across M0, M1, M2, and M3; P =0.66byKruskal-Wallis test across all groups. Females: P = 0.93 by ANOVA across F0, F1, F2, and F3; P = 0.92 by ANOVA across all groups. C:Males:P =0.37by ANOVA across M0, M1, M2, and M3; P = 0.86 by ANOVA across all groups. Females: P =0.76andP = 0.07 vs. 0F for 1F and 2F by posttest; P = 0.005 by ANOVA across all groups; P =0.29,P = 0.98, and P = 0.08 for F1, F2 and F3, respectively, compared with Con F by posttest. D:Males:P = 0.68 by ANOVA across all groups. Females: P = 0.70 by ANOVA across all groups. E:Males:P =0.36byKruskal-Wallistest across all groups. Females: P = 0.025 by ANOVA across all groups; P =0.30,P = 0.09, and P = 0.34 vs. F0 for F1, F2, and F3 by posttest. F:Males:P = 0.10 by ANOVA across all groups. Females: P = 0.19 by ANOVA across all groups.

Calculations mean value for each animal was calculated relative to the Historical Control Data total tissue area counted for each estimate. Volume frac- Data on pancreas weights were collected from the tions were measured in percentage of the total pancreas contract laboratories where the study of interest was volume. Total cell mass in milligrams was calculated by conducted to ensure matching for origin, age range, and multiplying the volume fraction with the total pancreas environmental conditions to the highest possible extent. weight. Historical control data on pancreas weights from animals used as vehicle controls in other studies of 4, 13, and 39– Statistics 52 weeks’ duration was used for comparison with liraglutide: The liraglutide data on pancreas weight were, per study 4 week (n = 29 males and 28 females as historical con- protocol, analyzed by ANCOVA on body weights and trols), 13 week (n = 19 males and 19 females as historic pancreas weights combining both sexes and using day controls), and 52 week (n = 22 males and 17 females as 0 body weight and terminal body weight, respectively, as fi historical controls). No historical control data were avail- covariates. These prespeci ed models were slightly differ- able for comparison with the 87-week study because this ent between studies because they were data-driven. Only is not a standard-length toxicology study. the 52-week study reported an increase in pancreas weight when this analysis was used (22). To obtain con- Quantitative Histology sistency across studies and thus enhance comparisons, Measurements were performed on samples obtained by post hoc modeling was performed using the same models a systematic uniform random sampling technique. The in all studies. In the prespecified analyses, terminal body diabetes.diabetesjournals.org Gotfredsen and Associates 2489 weight was taken into consideration in the statistical Recovery Animals. Additional animals were included to models. However, that is problematic because liraglutide explore reversibility of potential findings (n = two males lowers body weight. The post hocanalysisisthusaone- and two females, in the control and high-dose groups in way ANOVA of the pancreas weight for each study and the 13- and 52-week studies for liraglutide and semaglu- sex separately. In case of variance inhomogeneity, mea- tide). These additional animals were dosed for the same sured by means of the Bartlett and Brown-Forsythe duration as all other animals in the study but were kept tests, a Kruskal-Wallis test was performed to evaluate for additional 4 weeks after end of treatment before being overall group effects. Because statistically significant killed. In the 52 weeks’ study, there was no longer any effects of liraglutide were seen in some studies for difference in pancreas weight versus controls after the some of the doses, a one-way ANOVA was also per- 4-week treatment-free period (data not shown). formed for the relative pancreas weight to further exam- Pancreas Histology ine a possible effect. The Dunnett multiple comparison Histological examination of the pancreata from the 4-, test was used as posttest after ANOVA in cases where 13-, and 52-week studies is shown in Table 2. Represen- the overall group effect was significant. Parameters from tative histological sections of the different compartments quantitative histology were analyzed by the Student of pancreas after 52 or 87 weeks of dosing are illustrated t test for each sex, except for a-cell mass, where one- in Figs. 2 and 3. The histological examination did not way ANOVA was used. P , 0.05 was considered statis- reveal treatment-related differences between dosed and tically significant. Data are presented as mean 6 SEM control animals at any time point. The endocrine pancreas unless otherwise stated. revealed well-demarcated islets with normal pale islet cells after 52 and 87 weeks of dosing (Fig. 2). In the exocrine RESULTS pancreas, all ducts appeared normal, both the large main Liraglutide duct with the high columnar epithelium and abundant Pancreas Weight surrounding connective tissue (shown in Fig. 3, left side fi Comparison With In-Study Control Groups. Pancreatic panels, and in higher magni cation in Supplementary weight data from liraglutide studies are shown in Fig. 1. A Fig. 1) and the medium-sized interlobular ducts with the significant increase that was apparently not dose-related lower cuboidal epithelium and less surrounding connective was found in the 4-week study in male animals (P = 0.036 tissue (Fig. 3, right side panels). The small intercalated ducts fl by ANOVA), whereas no increase was seen in females in with at epithelium and no or sparse surrounding connec- fi the same study. No statistically significant differences tive tissue also appeared normal (not shown in gures). The were found for any dose level in male or female animals acinar cell parenchyma consisted of normal pyramid-shaped fi after 13 or 87 weeks of dosing. Similarly, in the 52-week cells where the apical part was lled with eosinophilic zymo- study, no statistically significant differences were found gen granules and the basophilic basal part contained the for the male animals but a significantly higher pancreas nucleus,asshowninFig.3andinSupplementaryFig.2. weight, apparently dose-related, was found in the female Quantitative Histology of Pancreas in 52-Week high-dose group compared with the in-study control Liraglutide Study group (P = 0.007 by ANOVA). A quantitative histological assessment of the pancreas in Comparison With Historical Control Data. This compar- the 52-week liraglutide study was conducted to evaluate ison did not show any statistically significant differences whether there were changes in the mass of pancreatic cell fi in pancreas weight for any dose level in males or females types that could not be identi ed by the qualitative b from the 4- and 13-week studies or males in the 52-week histological analysis. Table 1 shows absolute mass of -, a d study. When compared against historical controls in -, -, PP, duct, and acinar cells. Liraglutide-dosed mon- fi females from the 52-week study, a statistically signifi- keys showed no signi cant differences in any of these cant difference across groups was found (P =0.005by measures, except for absolute duct cell and acinar cell fi ANOVA), but posttest showed that only the in-study con- mass, which was signi cantly increased in the female trols exhibited a statistically significant different (lower) high-dose group. When the proportion of cells per volume pancreas weight than the historical controls (n =4and of pancreas was evaluated, no changes were found for 6 17, respectively; P =0.04). ductal cell volume in males (liraglutide 6.14 0.59 vs. control 6.71 6 0.62%, P = 0.53) or females (7.40 6 0.63 Comparison of Pancreas Weight Adjusted for Body vs. 6.55 6 1.29%, P = 0.57) or for acinar cell volume in Weight. An ANOVA analysis of the relative pancreas males (91.4 6 0.9 vs. 88.7 6 0.8%, P = 0.06) or females weight showed a significant increase compared with in- (90.3 6 0.4 vs. 90.0 6 0.7%, P = 0.71) high-dose com- study controls in the male mid- and high-dose groups and pared with vehicle. Thus, the increased pancreas weight in the female high-dose group in the 52-week liraglutide was a balanced increase of the exocrine pancreas, with no study (data not shown). There were no statistically apparent change in the ratio of ductal to acinar tissue. To significant differences in the 4-, 13-, or 87-week liraglu- further evaluate if a-cell mass specifically was changed by tide studies. liraglutide, the low- and medium-dose liraglutide groups 2490 GLP-1 Receptor Agonists and the Pancreas Diabetes Volume 63, July 2014

Figure 2—Liraglutide studies in nonhuman primates (hematoxylin and eosin staining). Endocrine pancreatic islets from males from 52 weeks’ study (left) and females from 87 weeks’ study (right) from control (upper row) or liraglutide high-dose group (lower row). Well- demarcated islets with normal-looking pale islet cells. Liraglutide-dosed animals look similar to what is seen in control animals. Original magnification 3200; bar indicates 100 mm.

were also evaluated quantitatively. Liraglutide did not significant treatment-related effect on pancreas weight change a-cell mass in males in the low- or middle-dose in males, whereas a significant difference was seen across groups (30.0 6 8.4 and 24.2 6 3.3, respectively, vs. groups in the females (P = 0.02 by ANOVA). However, by 29.9 6 3.6 mg for controls, from Table 1) or females posttest, no significant difference was found between (16.0 6 1.6 and 20.4 6 2.0, respectively, vs. 22.6 6 3.3 treated groups and the control group, and no apparent mg for controls, from Table 1), with P = 0.74 and P = 0.22 dose-related effects were seen. In the 52-week study, no for males and females across groups by ANOVA. statistically significant differences across study groups Examples of b- and non–b-cell staining (Fig. 4) and were observed for pancreas weight; the highest pancreas a-cell and proliferation (Ki-67) (Fig. 5, and higher magni- weights were found in the control groups. fication in Supplementary Fig. 3) in representative sections Histological examination of the pancreas from the of pancreata from males and females in the vehicle and the 13- and 52-week studies revealed common background liraglutide high-dose group of the 52-week dosing study are findings of minimal to mild severity and with a focal shown. In control and high-dose animals, glucagon staining distribution. There were no signs of treatment-related showed a high and variable number of a-cells in islets effects. The data are reported in Table 2. (typically ;50% of non–b-cells were a-cells). Small num- bers of single cells or small clusters of glucagon-positive DISCUSSION cells associated with ducts were seen in the exocrine pan- creas (Fig. 5). Very few cells in the endocrine and exocrine Reported here are further data from nonhuman primate pancreas were positive for Ki-67, and there was no appar- studies conducted with liraglutide as a supplement to ent difference between the liraglutide and control groups previously published data on pancreas histology in mice, (data not shown). As a positive control for proliferation, rats, and nonhuman primates (13). An apparent dose- lymph nodes present in 15 of the total of 32 pancreata related increase in absolute pancreas weight was found showed strong labeling of many cells in germinal centers in females in one of four monkey studies with liraglutide and also some single cells (shown as inserts in Fig. 5). and in none of two monkey studies with semaglutide, and an increase that did not appear to be dose-related was found Semaglutide in a 4-week liraglutide study, in males only (Table 3). A Pancreas weights from semaglutide studies are shown in comparison with historical control data was made. This is Fig. 1. In the 13-week study, there was no statistically a common way of setting toxicological data into perspective diabetes.diabetesjournals.org Gotfredsen and Associates 2491

Figure 3—Liraglutide studies in nonhuman primates (hematoxylin and eosin staining). Top row: 52 weeks’ study control males. Second row: 87 weeks’ study control females. Third row: 52 weeks’ study high-dose males. Bottom row: 87 weeks’ study high-dose females. Left: Original magnification 3200; bar indicates 100 mm. Ductular part of the exocrine pancreas with presence of a large main duct with columnar epithelium. The duct is surrounded by connective tissue but is still located within the parenchyma. The duct from high-dose group looks similar to that of the control monkeys. Right: Original magnification 3400; bar indicates 50 mm. Exocrine acinar cell paren- chyma and ductular part of the exocrine pancreas with presence of medium-sized interlobular ducts with cuboidal epithelium. The duct epithelium is lower and the amount of the surrounding connective tissue is lesser than for the large ducts. The duct from the high-dose group looks similar to that of the control monkeys. The acinar secretory compartment consists of pyramid-shaped cells where the apical part is filled with eosinophilic zymogen granules and the basophilic basal part contains the nucleus. The secretory compartment from high- dose group looks similar to that of the control monkeys. 2492 GLP-1 Receptor Agonists and the Pancreas Diabetes Volume 63, July 2014

Table 1—Absolute mass (mg) for endocrine cell types, duct, and acinar cells in pancreas from monkeys dosed with vehicle or liraglutide for 52 weeks Group (n =4) b-Cells (mg) a-Cells (mg) d-Cells (mg) PP cells (mg) Duct cells (mg) Acinar cells (mg) Male Vehicle control 49.1 6 4.4 29.9 6 3.6 17.3 6 2.5 3.73 6 1.06 251 6 36 3,319 6 454 Liraglutide 5 mg/kg/day 43.4 6 9.1 24.0 6 3.5 13.8 6 3.5 6.30 6 2.02 312 6 55 4,563 6 482 P 0.59 0.28 0.45 0.30 0.39 0.11 Female Vehicle control 45.4 6 10.3 22.6 6 3.3 10.1 6 3.1 3.32 6 2.26 199 6 53 2,694 6 450 Liraglutide 5 mg/kg/day 51.8 6 10.1 24.0 6 3.3 12.3 6 0.9 6.00 6 1.50 439 6 64 5,325 6 469 P 0.67 0.78 0.52 0.36 0.03 0.007 Data are shown as mean 6 SEM. P values compared with control group same sex. and may be especially useful in nonhuman primate studies 60-times the clinically relevant exposure for up to 87 weeks where the number of animals is low due to ethical reasons. or semaglutide dosing for up to 52 weeks in monkeys. The histological analysis of the pancreas from the liraglu- The studies reported here have a high relevance for tide and semaglutide studies did not reveal any potentially humans because the pancreata of nonhuman primates are adverse findings that could be related to treatment (e.g., closely related to humans anatomically and physiologi- pancreatitis, inflammatory cell infiltrations, or hyperpla- cally. Additionally, the morphology of nonhuman primate sia). Overall, this led to the conclusion that an increase islets is like that seen in humans and different from in pancreas weight cannot be ruled out, but no consis- rodents (23). Quantitative histology was used to assess tent dose-related increases in pancreas weight were seen changes in the mass of different pancreatic tissue compo- across the liraglutide and semaglutide studies in mon- nents, taking the three-dimensional structure of the or- keys. In combination with the lack of treatment-related gan into consideration (24). An inherent weakness of histopathological changes, the data showed no adverse these studies is the relatively limited number of non- effects on the pancreas by liraglutide dosing of up to human primates per group because number of animals is

Figure 4—Fifty-two weeks’ liraglutide study in nonhuman primates. Male (left) and female (right) from control (upper row) or liraglutide high- dose group (lower row). Double immunohistochemical staining for b-cells (reddish brown) and non–b-cells (the sum of glucagon, somato- statin, and PP, violet/black). Islet structure and distribution of b-, and non–b-cells from liraglutide-dosed animals look similar to what is seen in control animals. Original magnification 3400; bar indicates 100 mm. diabetes.diabetesjournals.org Gotfredsen and Associates 2493

Figure 5—Fifty-two weeks’ liraglutide study in nonhuman primates. Double immunohistochemical staining for glucagon (pink) and Ki-67 (black). Male (left) and female (right) from control (upper row) or liraglutide high-dose group (lower row). Glucagon staining shows a high and variable number of a-cells in islets and a small number of single cells and small clusters of glucagon-positive cells in the acinar exocrine pancreas and associated with ducts. The inserts show Ki-67 labeling in lymph nodes in the same sections. In total, 15 of 32 monkeys had lymph nodes in these double-stained sections but none of the high-dose treated male monkeys had lymph nodes in such sections. Original magnification 3200.

constrained for ethical reasons. However, this report is treatment. Quantitative histology of the pancreas from based on data from 90 and 48 animals dosed with liraglu- the 52-week study with liraglutide demonstrated an in- tide and semaglutide, respectively. Another limitation is crease of exocrine pancreas tissue with an apparently un- that only one transverse section was examined per ani- changed ratio between acinar and ductal cells and with mal. As a consequence thereof, the statistical power is normal tissue architecture. Cell proliferation in the pan- low. Three of the four studies that have described adverse creas measured by Ki-67 appeared unchanged by liraglu- effects of DPP-4is or GLP-1RAs on the pancreas were tide; very low proliferation rates were found in all performed in rodents (15,25,26). Thus, at the current animals. There appeared to be no change of b-, d-, or point in time with the very few studies available in non- PP cell mass by 52 weeks of liraglutide treatment in the human primates, the cumulative data in these studies high-dose group and there was apparently no change in have a strong relevance for the assessment of adverse a-cell mass or indication of proliferation of a-cells in any pancreas effects in humans, despite the relatively limited of the three liraglutide-dosed groups in the 52-week number of animals per group. study. The islets from the control and the liraglutide A recent study with human pancreata from patients groups in the 52-week study had ;50% non–b-endocrine previously treated with sitagliptin or exenatide (seven cells, with a substantial portion of those being a-cells (25– sitagliptin, one exenatide) reported an increased pancreas 30% of endocrine cell mass), less d-cells (;15%), and only weight (16). A number of important considerations with a few PP cells (4–7%). The fraction of the four endocrine the study design may have affected the results: the groups cell types in our study is in agreement with data from were unbalanced, with an 18-year age difference between human pancreata and a descriptive study in Cynomolgus the groups, and there was no attempt to control for type monkeys (31–34). The intraislet organization of the b- of diabetes, weight, age, or sex. Pancreas weight depends and non–b-cells was random, with no clear rodent-like on body weight, stage of diabetes, age, and sex (27,28). mantle and core, but with a more complex subunit struc- These methodological problems are clearly elucidated in ture of mantles and cores, as characteristic for nonhuman two related commentaries/reviews (29,30). primates and humans (23,35,36). In all groups, including The data presented here do not show an increase in cell the controls, the distribution of a-cells was identical in replication or number of a-cells caused by liraglutide islets and islets-like structures of variable size. The 2494 GLP-1 Receptor Agonists and the Pancreas Diabetes Volume 63, July 2014

Table 2—Summary of histopathological findings in the 4-, 13-, and 52-week liraglutide studies and in the 13- and 52-week semaglutide studies Males Females 4-week liraglutide dose levels (mg/kg/day) 0 0.05 0.5 5.0 0 0.05 0.5 5.0 Animals examined 3 3 3 3 3 3 3 3 Endocrine pancreas No abnormality detected 3 3 3 3 3 3 3 3 Exocrine pancreas No abnormality detected 3 3 3 3 3 3 3 3 13-week liraglutide dose levels (mg/kg/day) 0 0.05 0.5 5.0 0 0.05 0.5 5.0 Animals examined 4 4 4 4 4 4 4 4 Endocrine pancreas No abnormality detected 4 4 3 3 2 3 4 4 Prominent islets 0 0 1 1 1 0 0 0 Fat infiltration, minimal 0 0 0 0 1 1 0 0 Exocrine pancreas No abnormality detected 4 4 4 4 3 4 4 4 Inflammatory cell infiltration, minimal 0 0 0 0 1 0 0 0 52-week liraglutide dose levels (mg/kg/day) 0 0.05 0.50 5.0 0 0.05 0.5 5.0 Animals examined 4 4 4 4 4 4 4 4 Endocrine pancreas No abnormality detected 4 4 4 4 4 4 4 4 Exocrine pancreas No abnormality detected 4 3 4 3 4 3 3 4 Fibrosis, focal, minimal 0 1 0 0 0 0 0 0 Inflammatory cell infiltration, focal, minimal 0 0 0 1 0 1 0 0 Ectopic splenic tissue 0 0 0 0 0 0 1 0 13-week semaglutide dose levels (mg/kg twice weekly) 0 0.004 0.086 0.47 0 0.004 0.086 0.47 Animals examined 4 4 4 4 4 4 4 4* Endocrine pancreas No abnormality detected 4 4 4 4 4 4 4 3 Islet atrophy 0 0 0 0 0 0 0 1 Exocrine pancreas No abnormality detected 4 4 4 3 3 1 4 3 Chronic focal inflammation, mild 0 0 0 0 0 1 0 0 Inflammatory cell foci, minimal 0 0 0 1 1 1 0 1 Focal brown pigment, minimal 0 0 0 0 0 1 0 0 Ectopic splenic tissue 0 0 0 0 0 1 0 0 52-week semaglutide dose levels (mg/kg twice weekly) 0 0.01 0.06 0.36 0 0.01 0.06 0.36 Animals examined 4 4 4 4 4 4 4 4 Endocrine pancreas No abnormality detected 4 4 4 4 4 4 4 4 Exocrine pancreas No abnormality detected 3 3 4 4 3 1 3 3 Focal arteritis/periarteritis, minimal 0 0 0 0 0 1 0 0 Focal interstitial inflammatory cell infiltration, minimal 1 1 0 0 0 2 1 1 Focal periductal inflammatory cell infiltration, slight 0 0 0 0 1 0 0 0 *Two animals were killed after 36 days, see Fig. 1 legend for details.

control and liraglutide-dosed groups both showed a num- studies with liraglutide. Despite this apparently dose- ber of single glucagon-positive cells in the exocrine area, related increase in pancreas weight, no histopathology single cells and small clusters of glucagon-positive cells was associated thereto, and therewerenopancreatic associated to the epithelial lining of both main and intraepithelial neoplasia lesions (13). smaller ducts. This finding of small clusters and single The four studies that have suggested adverse effect of endocrine cells is normal in Cynomolgus monkeys (34). A DPP-4is or GLP-1RAs on the pancreas have described or similar pattern with single cells in the exocrine areas and discussed increased risk for pancreatitis, metaplasia, or associated to ductal structures was observed for b-cells inflammation and pancreatic adenocarcinomas and gluca- and less frequently with d-cells and PP cells, with no gonomas (15,16,25,26). In contrast, hundreds of other differences between groups. An apparently dose-related studies have investigated effects of these drugs on the increase in absolute pancreas weight was found in one pancreas but have not reported adverse effects; a few of four studies, and only in one sex, in nonhuman primate are referenced here for liraglutide (37–39). A study, diabetes.diabetesjournals.org Gotfredsen and Associates 2495

Table 3—Summary of dose-related tendencies for absolute pancreas weight (g) across studies with liraglutide and semaglutide Pancreas weight, tendency for dose-related change? Males Females Consistent in both sexes 4-week liraglutide * No 13-week liraglutide Yes 52-week liraglutide ** No 87-week liraglutide Yes 13-week semaglutide Yes 52-week semaglutide Yes *P , 0.05 for a statistically significant increase in absolute pancreas weight, but not apparently dose-related. **P , 0.01 for an apparently dose-related increase in absolute pancreas weight. performed after others had reported potential adverse study design to liraglutide or placebo and will be followed findings, confirmed the absence of pathology in diabetic up for a minimum of 42 months for the primary end point rats and also did not show any regional differences in the of adjudicated macrovascular events, including nonfatal pancreas induced by liraglutide when the pancreas was myocardial infarction, stroke, or cardiovascular death. Ad- divided into four regions and examined by stereology judication of all adverse reactions related to pancreatitis (40). A recent publication used a human islet amyloid and any neoplasm is an integral part of the protocol polypeptide transgenic model similar to one of the earlier throughout the duration of the LEADER study. Random- studies, just in mice instead of in rats, treats the animals ized, controlled, long-duration trials with independent ad- for 1 year instead of 12 weeks, used 20–25 animals in judication are the only way to evaluate rare side effects, as each group instead of 8, and found no pancreas pathology also recently mentioned by Kahn (29) and Drucker (30). associated with sitagliptin (25,41). The U.S. Food and The LEADER study will report in 2016. Drug Administration recently published that it reassessed data from 50 GLP-1 based therapeutics and found no changes indicating pancreatic injury (42). To understand Acknowledgments. The authors thank Charles Pyke, Novo Nordisk, for careful reading of the manuscript and interpretations of the methodology used. whether there are any potential adverse effects of GLP- Duality of Interest. Novo Nordisk markets liraglutide for the treatment of 1RAs on the human pancreas, the expression pattern of diabetes and has semaglutide in phase 3 clinical development. All authors are GLP-1R may be important. It has recently been recog- full-time employees of Novo Nordisk and hold minor share portions as part of nized that most studies measuring the GLP-1R may be their employment. No other potential conflicts of interest relevant to this article invalid because the antibodies used are not specific for were reported. GLP-1R (30,43,44). G-protein–coupled receptors are no- Author Contributions. C.F.G. performed the quantitative histology. toriously known for this problem (45–47). Some scientific A.-M.M. collected and reviewed parts of the data and took part in drawing journals have provided new guidance for validation conclusions from the liraglutide studies. I.T. reviewed the histopathological data experiments that must be available for reliable documen- and selected the photographs. N.C.B.N. took part in drawing conclusions from the tation of expression of a G-protein–coupled receptor (48). liraglutide studies. Z.S. designed the semaglutide studies and concluded the However, valid studies are available documenting GLP-1R results for those. L.B.K. took part in drawing conclusions from the results of the studies and wrote the major part of ABSTRACT, INTRODUCTION, and DISCUSSION expression (49). These studies measure receptor expres- sections. M.O.L. wrote the major part of the RESEARCH DESIGN AND METHODS and RESULTS sion by ligand binding and show that pancreatic adeno- sections and some parts of the ABSTRACT and DISCUSSION and performed the statis- carcinomas do not express GLP-1R (49,50). Thus, from tical analysis. All authors reviewed the manuscript. L.B.K. is the guarantor of this a molecular target point of view, it seems unlikely that work and, as such, had full access to all the data in the study and takes re- GLP-1R agonism should directly worsen or induce pancre- sponsibility for the integrity of the data and the accuracy of the data analysis. atic adenocarcinomas when such tumors do not express GLP-1R. References On the basis of the totality of information available to 1. Doyle ME, Egan JM. Mechanisms of action of glucagon-like peptide 1 in the Novo Nordisk, the available information is insufficient to pancreas. Pharmacol Ther 2007;113:546–593 confirm or exclude an association between liraglutide and 2. Campbell JE, Drucker DJ. Pharmacology, physiology, and mechanisms of – pancreatitis, and there is no evidence that it increases the incretin hormone action. Cell Metabolism 2013;17:819 837 risk of pancreatic cancer in patients with type 2 diabetes. 3. Lamont BJ, Li Y, Kwan E, Brown TJ, Gaisano H, Drucker DJ. Pancreatic GLP-1 receptor activation is sufficient for incretin control of glucose metabolism The Liraglutide Effect and Action in Diabetes: Evaluation – — in mice. J Clin Invest 2012;122:388 402 of Cardiovascular Outcome Results A Long Term Evalu- 4. Ali S, Lamont BJ, Charron MJ, Drucker DJ. Dual elimination of the glucagon ation (LEADER; NCT01179048) will prospectively evalu- and GLP-1 receptors in mice reveals plasticity in the incretin axis. J Clin Invest ate the overall safety of liraglutide. The trial has enrolled 2011;121:1917–1929 9,340 patients with type 2 diabetes and a high cardiovascu- 5. Mentlein R. Dipeptidyl-peptidase IV (CD26)—role in the inactivation of lar risk profile. Patients are randomized 1:1 in a double-blind regulatory peptides. Regul Pept 1999;85:9–24 2496 GLP-1 Receptor Agonists and the Pancreas Diabetes Volume 63, July 2014

6. Herman GA, Bergman A, Stevens C, et al. Effect of single oral doses of 24. Boyce JT, Boyce RW, Gundersen HJ. Choice of morphometric methods sitagliptin, a dipeptidyl peptidase-4 inhibitor, on incretin and plasma glucose and consequences in the regulatory environment. Toxicol Pathol 2010;38: levels after an oral glucose tolerance test in patients with type 2 diabetes. J Clin 1128–1133 Endocrinol Metab 2006;91:4612–4619 25. Matveyenko AV, Dry S, Cox HI, et al. Beneficial endocrine but adverse 7. Aaboe K, Knop FK, Vilsbøll T, et al. Twelve weeks treatment with the DPP-4 exocrine effects of sitagliptin in the human islet amyloid polypeptide transgenic inhibitor, sitagliptin, prevents degradation of peptide YY and improves glucose rat model of type 2 diabetes: interactions with metformin. Diabetes 2009;58: and non-glucose induced insulin secretion in patients with type 2 diabetes 1604–1615 mellitus. Diabetes Obes Metab 2010;12:323–333 26. Nachnani JS, Bulchandani DG, Nookala A, et al. Biochemical and histo- 8. Buse JB, Garber A, Rosenstock J, et al. Liraglutide treatment is associated logical effects of exendin-4 (exenatide) on the rat pancreas. Diabetologia 2010; with a low frequency and magnitude of antibody formation with no apparent 53:153–159 impact on glycemic response or increased frequency of adverse events: results 27. Williams AJK, Thrower SL, Sequeiros IM, et al. Pancreatic volume is re- from the Liraglutide Effect and Action in Diabetes (LEAD) trials. J Clin Endocrinol duced in adult patients with recently diagnosed type 1 diabetes. J Clin Endocrinol Metab 2011;96:1695–1702 Metab 2012;97:E2109–E2113 9. Fineman MS, Mace KF, Diamant M, et al. Clinical relevance of anti- 28. Cuntz U, Frank G, Lehnert P, Fichter M. Interrelationships between the size exenatide antibodies: safety, efficacy and cross-reactivity with long-term treat- of the pancreas and the weight of patients with eating disorders. Int J Eat Disord ment. Diabetes Obes Metab 2012;14:546–554 2000;27:297–303 10. Rosenstock J, Balas B, Charbonnel B, et al.; T-emerge 2 Study Group. The 29. Kahn SE. Incretin therapy and islet pathology: a time for caution. Diabetes fate of taspoglutide, a weekly GLP-1 receptor agonist, versus twice-daily exenatide 2013;62:2178–2180 for type 2 diabetes: the T-emerge 2 trial. Diabetes Care 2013;36:498–504 30. Drucker DJ. Incretin action in the pancreas: potential promise, possible 11. Nauck MA, Petri JR, Sesti G, et al. The once-weekly human GLP-1 analogue perils, and pathological pitfalls. Diabetes 2013;62:3316–3323 semaglutide provides significant reductions in HbA1c and body weight in patients 31. Stefan Y, Orci L, Malaisse-Lagae F, Perrelet A, Patel Y, Unger RH. Quan- with type 2 diabetes. Abstracts of the 48th EASD (European Association for the titation of endocrine cell content in the pancreas of nondiabetic and diabetic Study of Diabetes) Annual Meeting of the European Association for the Study of humans. Diabetes 1982;31:694–700 Diabetes. October 1-5, 2012. Berlin, Germany. Diabetologia 2012;55(Suppl.):S7 32. Rahier J, Goebbels RM, Henquin JC. Cellular composition of the human 12. World Health Organization. WHO guidelines on the quality, safety, and diabetic pancreas. Diabetologia 1983;24:366–371 efficacy of biotherapeutic products prepared by recombinant DNA technol- 33. Street CN, Lakey JR, Shapiro AM, et al. Islet graft assessment in the Ed- ogy. Available from http://www.who.int/biologicals/WHO_rDNA_2nd_public_ monton Protocol: implications for predicting long-term clinical outcome. Diabetes consultation_28_June_2013.pdf. Accessed 14 Oct 2013 2004;53:3107–3114 13. Nyborg NC, Mølck AM, Madsen LW, Knudsen LB. The human GLP-1 analog 34. Wieczorek G, Pospischil A, Perentes E. A comparative immunohistochemical liraglutide and the pancreas: evidence for the absence of structural pancreatic study of pancreatic islets in laboratory animals (rats, dogs, minipigs, nonhuman changes in three species. Diabetes 2012;61:1243–1249 primates). Exp Toxicol Pathol 1998;50:151–172 14. Koehler JA, Baggio LL, Lamont BJ, Ali S, Drucker DJ. Glucagon-like 35. Kim A, Miller K, Jo J, Kilimnik G, Wojcik P, Hara M. Islet architecture: peptide-1 receptor activation modulates pancreatitis-associated gene expression A comparative study. Islets 2009;1:129–136 but does not modify the susceptibility to experimental pancreatitis in mice. Di- 36. Steiner DJ, Kim A, Miller K, Hara M. Pancreatic islet plasticity: interspecies abetes 2009;58:2148–2161 comparison of islet architecture and composition. Islets 2010;2:135–145 15. Gier B, Matveyenko AV, Kirakossian D, Dawson D, Dry SM, Butler PC. 37. Bock T, Pakkenberg B, Buschard K. The endocrine pancreas in non-diabetic Chronic GLP-1 receptor activation by exendin-4 induces expansion of pancreatic rats after short-term and long-term treatment with the long-acting GLP-1 de- duct glands in rats and accelerates formation of dysplastic lesions and chronic rivative NN2211. APMIS 2003;111:1117–1124 pancreatitis in the Kras(G12D) mouse model. Diabetes 2012;61:1250–1262 38. Raun K, von Voss P, Gotfredsen CF, Golozoubova V, Rolin B, Knudsen LB. 16. Butler AE, Campbell-Thompson M, Gurlo T, Dawson DW, Atkinson M, Butler Liraglutide, a long-acting glucagon-like peptide-1 analog, reduces body weight PC. Marked expansion of exocrine and endocrine pancreas with incretin therapy and food intake in obese candy-fed rats, whereas a dipeptidyl peptidase-IV in- in humans with increased exocrine pancreas dysplasia and the potential for hibitor, vildagliptin, does not. Diabetes 2007;56:8–15 glucagon-producing neuroendocrine tumors. Diabetes 2013;62:2595–2604. 39. Cummings BP, Stanhope KL, Graham JL, et al. Chronic administration of the 17. Ruehl-Fehlert C, Kittel B, Morawietz G, et al.; RITA Group; NACAD Group. glucagon-like peptide-1 analog, liraglutide, delays the onset of diabetes and Revised guides for organ sampling and trimming in rats and mice—part 1. Exp lowers triglycerides in UCD-T2DM rats. Diabetes 2010;59:2653–2661 Toxicol Pathol 2003;55:91–106 40. Vrang N, Jelsing J, Simonsen L, et al. The effects of 13 weeks of liraglutide 18. Crissman JW, Goodman DG, Hildebrandt PK, et al. Best practices guideline: treatment on endocrine and exocrine pancreas in male and female ZDF-rats: toxicologic histopathology. Toxicol Pathol 2004;32:126–131 a quantitative and qualitative analysis revealing no evidence of drug-induced 19. Bock T, Svenstrup K, Pakkenberg B, Buschard K. Unbiased estimation of pancreatitis. Am J Physiol Endocrinol Metab 2012;303:E253–264 total beta-cell number and mean beta-cell volume in rodent pancreas. APMIS 41. Aston-Mourney K, Subramanian SL, Zraika S, et al. One year of sitagliptin 1999;107:791–799 treatment protects against islet amyloid-associated b-cell loss and does not 20. Gundersen HJ. The smooth fractionator. J Microsc 2002;207:191–210 induce pancreatitis or pancreatic neoplasia in mice. Am J Physiol Endocrinol 21. Larsen MO, Wilken M, Gotfredsen CF, Carr RD, Svendsen O, Rolin B. Mild Metab 2013;305:E475–E484 streptozotocin diabetes in the Göttingen minipig. A novel model of moderate 42. Hummer B. FDA surveillance of adverse drug effects (Abstract). NIH Pan- insulin deficiency and diabetes. Am J Physiol Endocrinol Metab 2002;282: creatitis, Diabetes, Pancreatic Cancer Workshop, Bethesda, MD, 12–13 June E1342–E1351 2013:8-9 22. Novo Nordisk: Liraglutide (injection) for the treatment of patients with 43. Panjwani N, Mulvihill EE, Longuet C, et al. GLP-1 receptor activation in- type 2 diabetes. NDA 22-341. Available from http://www.fda.gov/downloads/ directly reduces hepatic lipid accumulation but does not attenuate development advisorycommittees/committeesmeetingmaterials/drugs/endocrinologicand of atherosclerosis in diabetic male ApoE(-/-) mice. Endocrinology 2013;154:127– metabolicdrugsadvisorycommittee/ucm148659.pdf. Accessed 11 July 2013 139 23. Kharouta M, Miller K, Kim A, et al. No mantle formation in rodent islets—the 44. Pyke C, Knudsen LB. The glucagon-like peptide-1 receptor—or not? prototype of islet revisited. Diabetes Res Clin Pract 2009;85:252–257 Endocrinology 2013;154:4–8 diabetes.diabetesjournals.org Gotfredsen and Associates 2497

45. Michel MC, Wieland T, Tsujimoto G. How reliable are G-protein-coupled re- 48. Gore AC. Editorial: antibody validation requirements for articles published in ceptor antibodies? Naunyn Schmiedebergs Arch Pharmacol 2009;379:385–388 endocrinology. Endocrinology 2013;154:579–580 46. Pradidarcheep W, Stallen J, Labruyère WT, Dabhoiwala NF, Michel MC, Lamers 49. Körner M, Stöckli M, Waser B, Reubi JC. GLP-1 receptor expression in WH. Lack of specificity of commercially available antisera against muscarinergic and human tumors and human normal tissues: potential for in vivo targeting. J Nucl adrenergic receptors. Naunyn Schmiedebergs Arch Pharmacol 2009;379:397–402 Med 2007;48:736–743 47. Hamdani N, van der Velden J. Lack of specificity of antibodies directed 50. Wild D, Christ E, Caplin ME, et al. Glucagon-like peptide-1 versus so- against human beta-adrenergic receptors. Naunyn Schmiedebergs Arch Phar- matostatin receptor targeting reveals 2 distinct forms of malignant insulinomas. macol 2009;379:403–407 J Nucl Med 2011;52:1073–1078