The Intestinotrophic Effects of Glucagon-Like Peptide-2 in Relation to Intestinal Neoplasia

Abstract

Context

Glucagon-like peptide-2 (GLP-2) is a gastrointestinal hormone with intestinotrophic and antiapoptotic effects. The hormone’s therapeutic potential in intestinal diseases and relation to intestinal neoplasia has raised great interest among researchers. This article reviews and discusses published experimental and clinical studies concerning the growth-stimulating and antiapoptotic effects of GLP-2 in relation to intestinal neoplasia.

Evidence Acquisition

The data used in this narrative review were collected through literature research in PubMed using English keywords. All studies to date examining GLP-2’s relation to intestinal neoplasms have been reviewed in this article, as the studies on the matter are sparse.

Evidence Synthesis

GLP-2 has been found to stimulate intestinal growth through secondary mediators and through the involvement of Akt phosphorylation. Studies on rodents have shown that exogenously administered GLP-2 increases the growth and incidence of adenomas in the colon, suggesting that GLP-2 may play an important role in the progression of intestinal tumors. Clinical studies have found that exogenous GLP-2 treatment is well tolerated for up to 30 months, but the tolerability for even longer periods of treatment has not been examined.

Conclusion

Exogenous GLP-2 is currently available as teduglutide for the treatment of short bowel syndrome. However, the association between exogenous GLP-2 treatment and intestinal neoplasia in humans has not been fully identified. This leads to a cause for concern regarding the later risk of the development or progression of intestinal tumors with long-term GLP-2 treatment. Therefore, further research regarding GLP-2’s potential relation to intestinal cancers is needed.

Glucagon-like peptide-2 (GLP-2) is a hormone with various effects. Particularly, the growth-stimulating effects of GLP-2 are of interest, as the hormone is known to increase the length of the crypt-villus axis of intestinal mucosa (1), suppress apoptosis of intestinal cells (2, 3), and improve the adaptive response to intestinal resection (4–6). In addition, GLP-2 has been found to increase blood flow to mesenteric vessels (7–10), decrease gastrointestinal motility (1), and reduce mucosal damage caused by inflammation (11). These findings have directed research toward the therapeutic benefits of GLP-2 in the treatment of gastrointestinal diseases, such as short bowel syndrome (SBS), inflammatory bowel disease, and chemotherapy-induced mucositis (6, 12–16); the findings have also raised concerns regarding the hormone’s potential association to intestinal cancers. This article aims to review the existing data regarding the intestinotrophic effects of GLP-2 and its relation to intestinal neoplasia.

Synthesis, Secretion, and Metabolism of GLP-2

GLP-2 is a 33–amino acid peptide hormone secreted by enteroendocrine l-cells together with glucagon-like peptide-1 (GLP-1) (2, 17, 18). The two hormones are produced through the proteolytic cleavage of proglucagon (2, 18). The release of GLP-2 is stimulated by meal ingestion, and its secretion follows nutrient-dependent behavior under normal physiological conditions (18, 19).

The enteroendocrine l-cells, responsible for the secretion of GLP-1 and GLP-2, are mainly localized in the distal ileum and throughout the colon (20–23).

Once secreted, endogenous GLP-2 (1–33) is degraded to its metabolite GLP-2 (3–33) by the enzyme dipeptidyl peptidase-4 (DPP-4) (18, 24, 25), yielding the metabolite GLP-2 (3–33) (18).

Inhibition of DPP-4 and Its Influence on the Proliferative Effect of GLP-2

The half-life (t_1/2) of GLP-2 (1–33) is approximately 5 to 7 minutes (17, 18), depending on the renal clearance and the enzymatic activity of DPP-4 (18, 26, 27).

Masur et al. (28) examined how inhibition of the enzyme DPP-4 affected the proliferative actions of GLP-2 in two different types of human colon cell lines. The results revealed that GLP-2 stimulation and simultaneous DPP-4 inhibition led to increased proliferation and migratory activity of both cell lines (28). Based on these data, Masur et al. (28) suggested that DPP-4 inhibitors might increase the risk of tumor growth enhancement and metastasis. Because DPP-4 inhibitors increase not only the half-life of GLP-1 but also that of GLP-2, the study presented a cause for concern regarding the potential tumor growth in patients treated with DPP-4 inhibitors and undiagnosed adenomas of the colon (28).

To support this result, Okawada et al. (29) found improved intestinal adaptation, increased intestinal proliferation, and increased GLP-2 secretion in mice with SBS treated with the DPP-4 inhibitor sitagliptin, and Hartmann et al. (30) also found a noticeable enhancement in the intestinotrophic effects of GLP-2 when DPP-4 was inhibited in mice and rats with the unspecific DPP-4 inhibitor valine pyrrolidide. However, to our knowledge, the relation among GLP-2 treatment, DPP-4 inhibition, and colon cancer has never been examined or discussed in these studies.

In contrast, only a modest increase in the small intestinal weight and no difference in the morphology were found in another study, in which long-term treatment with sitagliptin in relation to intestinal neoplasia was investigated in mice (31). Furthermore, this study found no signs of the influence of sitagliptin on intestinal neoplasia (31).

Whether DPP-4 inhibition promotes intestinal growth by increasing the level of GLP-2 is not clearly identified, but the issue represents an interesting topic for future research.

The GLP-2 Receptor

The GLP-2 receptor (GLP-2R) is expressed in the gastrointestinal tract and central nervous system (32). The receptor was cloned from both humans and rats in 1999 by Munroe et al. (32). Using fluorescence in situ hybridization, the human GLP-2R gene was localized to chromosome 17p13.3 (32).

GLP-2R is characterized as a 7–transmembrane domain G protein–coupled receptor from the B glucagon-secretin subfamily of G protein–coupled receptors (26, 32). The receptor for GLP-2 shows great similarity to the receptors for glucagon and GLP-1 (32). However, the GLP-2 receptor is highly specific to GLP-2 alone and does not recognize related members of the glucagon peptide family (32–36).

Studies using different methods of in situ hybridization and antisera have revealed variable results regarding the cellular localization of GLP-2R (32, 37–39). GLP-2R is localized to murine enteric neurons (37), enteroendocrine cells in the gut (38), and intestinal myofibroblasts (39). The cellular distribution of GLP-2R is thoroughly discussed, and further research regarding the exact localization of GLP-2R is required. However, GLP-2R expression in enterocytes or crypt cells has never been identified (40), suggesting that the observed proliferative and antiapoptotic effects of GLP-2 in the gastrointestinal tract must be indirectly induced.

GLP-2R Expression in Gastrointestinal Tumors

Following the abovementioned studies, Körner et al. (41) examined GLP-2R expression in different gastrointestinal tumors. The team extracted tumors from different gastrointestinal origins (such as colorectal adenocarcinomas, gastric adenocarcinomas, and gastrointestinal stromal tumors) and extragastrointestinal origins (such as small cell lung cancers, rhabdomyosarcomas, and leiomyosarcomas) and compared the GLP-2R expression of these samples with normal tissue. Samples from nonneoplastic gastrointestinal diseases, such as Crohn disease, ulcerative colitis, and Hirschsprung disease, were also used in the study (41).

The study confirmed a notable expression of GLP-2R messenger RNA (mRNA) in gastrointestinal stromal tumors, as 15 of 22 of the analyzed gastrointestinal stromal tumor samples demonstrated GLP-2R mRNA expression (41). GLP-2R mRNA was also expressed in four of four ileal carcinoid tumors and in two of seven colon adenocarcinomas. However, no GLP-2R expression was found in adenocarcinomas when using in vitro autoradiography (41). That GISTs showed GLP-2R expression was of particular notice (38, 41).

In a similar study on human colorectal cancers and polyps, only 20% of the adenocarcinoma samples showed expression of GLP-2R (42). Nonmalignant samples did not stain positive when examined with immunohistochemistry (IHC) (42). For comparison, the enteroendocrine cells from normal colon mucosa all showed immune activity for GLP-2R (42).

In contrast to the findings of the aforementioned study, this study suggested that the expression of GLP-2R might not play a role in colon adenoma pathogenesis (42). This seems interesting, because the study by Körner et al. (41) found GLP-2R expression in ileal carcinoid tumors and colon adenocarcinomas using RT-PCR but not autoradiography, suggesting a significance in the choice of method used.

Intracellular and Extracellular Pathways Activated by GLP-2

Studies have shown that GLP-2 enhanced intestinal IGF-I expression and secretion and have concluded that IGF-I may be an important mediator of the trophic effects of GLP-2 in the gut (3, 43, 44). Because GLP-2 also stimulates the enhancement of gut-barrier function (45), studies have further examined and discussed whether this is due to activation of β-catenin signaling through IGF-I, as β-catenin regulates cell-to-cell adhesion among other regulatory effects (46–48). The involvement of IGF-I and β-catenin in the GLP-2–induced effects on the small and large bowel is of particular interest, because IGF-I is a growth factor, and β-catenin is a known proto-oncogene that is often overactivated or dysregulated in many types of cancers, particularly colorectal cancers (43, 46, 49, 50). Research has found that IGF-I activates the β-catenin pathway through inhibition of glycogen synthase kinase-3β (GSK-3β) and activation of downstream Ras in hepatoma cells (49, 51). Inhibition of GSK-3β is known to result in cytoplasmic accumulation and nuclear translocation of β-catenin (46, 49, 52, 53), raising a concern regarding the link between GLP-2–induced IGF-I secretion and GSK-3β inhibition. This concern has led to questions regarding the involvement of IGF-I in the growth-stimulating effects of GLP-2 and whether GLP-2 may promote cytoplasmic accumulation of β-catenin through increased IGF-I secretion.

Dubé et al. (54) studied the relationship among GLP-2, IGF-I and β-catenin. The study examined whether activation of the Wnt/β-catenin signaling pathway and phosphatidylinositol 3-kinase (PI3K)/Akt pathway was mediated through GLP-2–induced IGF-I secretion, as a previous study had demonstrated that GLP-2 dose dependently activated Akt signaling in the neonatal pig intestine (55). Because Akt phosphorylation also increases the transcriptional activity of β-catenin (56), the study especially focused on whether GLP-2–induced Akt phosphorylation strictly depends on IGF-I (54).

Using IHC, Western blotting, and RT-PCR, the authors showed that GLP-2 caused nuclear translocation of β-catenin in some but not all of the proliferative crypt cells and that GLP-2 increased c-myc mRNA expression (54). The study also found that the elimination of IGF-IR/IGF-I signaling resulted in decreased β-catenin translocation and decreased GLP-2–induced epithelial proliferation in the mouse intestine (54). However, the study emphasized that IGF-IR/IGF-I signaling was only required for the GLP-2–induced alterations in β-catenin. The IGF-IR/IGF-I signaling pathway was not required to activate GLP-2–mediated Akt phosphorylation in the murine intestine, as Akt phosphorylation was preserved in Igf1 knockout mice (54). Therefore, Dubé et al. (54) concluded that GLP-2 activates Akt signaling in the intestinal mucosa, crypt, and villus epithelium in mice, but this activation is not strictly dependent on IGF-IR/IGF-I signaling. They suggested that other indirect mediators could be responsible for GLP-2–induced Akt phosphorylation (54).

In a subsequent study, a correlation between ErbB signaling and GLP-2–induced stimulation of bowel growth was reported (40, 57). The ErbB signaling network consists of different epidermal growth factors that function as ligands to corresponding epidermal growth factor receptors [(EGFRs)/ErbBs] (57–59). The binding of a ligand to its matching receptor causes an autophosphorylation of the receptor and, thereby, activation of downstream kinases, such as Akt and MAPK (40, 58, 60). Akt is known for its regulation of the cell cycle, cell survival, cell growth, and cell metabolism (61–64). The study demonstrated that exogenously administered GLP-2 promoted the expression of a subset of ErbB ligands with known relations to cancer, such as amphiregulin (65), epiregulin (66, 67), and heparin binding–epidermal growth factor (68), whereas IGF-I and KGF were found to have no effect on the expression of the aforementioned ErbB ligands (40). Interestingly, the study also concluded that the induction of ErbB ligands by GLP-2 required the presence of GLP-2R, as the findings were not detected in Glp2r knockout mice (40).

Further studies have subsequently revealed that GLP-2R activation results in the initiation of the PI3K-dependent Akt-mTOR signaling pathway (69, 70). Using GLP-2R–expressing transfected HEK 293 cells, Shi et al. (70) found that the PI3K-dependent Akt-mTOR signaling pathway is the main GLP-2–activated intracellular pathway. The results of the study showed that GLP-2, possibly through unknown secondary mediators, dose-dependently phosphorylated Akt and that inhibition of PI3K blocked this phosphorylation, suggesting a correlation between GLP-2 and PI3K-dependent Akt signaling (70). Other in vivo studies in animals have also indicated that stimulation of cellular growth by GLP-2 in the intestine is associated with activation of Akt signaling and IGF-IR/IGF-I signaling (69, 71).

Despite these findings, it remains uncertain how GLP-2 activates the Akt signaling pathway (70). Growth factors and hormones are known to stimulate PI3K activity and Akt signaling through PI3K (52, 61, 62, 72, 73). However, how GLP-2 specifically induces the phosphorylation of Akt remains unknown. The connection between Akt signaling and GLP-2 stimulation has led to the speculation that secondary growth factors, such as IGF-I and ErbB ligands, may be indirectly responsible for GLP-2–induced intestinal growth (40, 70).

The studies discussed suggest different mediators to be involved in GLP-2–induced intestinal growth. One observation, however, seems to be mutual in the analyzed studies; all the studies found a connection between GLP-2R stimulation and intracellular Akt phosphorylation (3, 40, 54, 69–71). This seems interesting because dysregulation of the PI3K/Akt pathway is often seen in most human tumors (73–76). Furthermore, Akt is a well-known proto-oncogene overexpressed in most human cancers because the kinase is responsible for the regulation of cell cycle progression, cell survival, and cell proliferation (52, 61, 63, 72, 73). If GLP-2 indeed increases the expression and secretion of IGF-1 and upregulates the transcription of ErbB ligands, leading to Akt phosphorylation and nuclear relocation of β-catenin in intestinal cells, there is some reason for concern because these abovementioned cellular effects are strongly associated with cancer and tumor growth (77, 78).

Table 1 summarizes the molecular effects of GLP-2 and the hormone’s mechanisms of action in relation to tumorigenesis.

Table 1.

Results of the Examined Studies on the Molecular Effects of GLP-2 in Relation to SBS and Tumorigenesis

Signaling Pathway	GLP-2–Related Mechanism of Action	SBS-Related Mechanisms	Tumorigenesis-Related Mechanisms
IGF-I/IGF-IR	GLP-2 enhances the intestinal expression and secretion of IGF-I, causing proliferation of the intestinal mucosa (3).	IGF-I is considered the main mediator of the intestinotrophic effects of GLP-2 (3, 43). Proliferation of the intestinal mucosa in response to GLP-2 treatment reduces the need for parenteral nutrition (16, 79).	GLP-2 induces epithelial proliferation in the mouse intestine through IGF-I/IGF-IR signaling (3). Exogenous GLP-2 increased growth of adenomas in the murine intestine (31, 80–82).
ErbB/EGFR	GLP-2 increases the expression of the ErbB/EGFR ligands: amphiregulin, epiregulin, and heparin binding–epidermal growth factor (69).	EGFR signaling plays an important role in the small bowel adaptation in patients with SBS; hence, activation of the signaling pathway promotes growth of the intestine (83).	ErbB/EGFR ligands have a known relation to cancer, as they activate downstream kinases such as Akt and MAPK (58, 65–68).
β-catenin	GLP-2 actives β-catenin signaling in the small intestinal crypt cells and causes nuclear translocation of β-catenin (54).	GLP-2 induces enhancement of gut barrier function presumably through β-catenin signaling as β-catenin regulates cell-to-cell adhesion (46, 54, 78).	β-catenin is a known proto-oncogene that is often overactivated or deregulated in many types of cancers (50, 53, 56, 78). GLP-2–treated mice had a higher expression of β-catenin (81).
Akt	GLP-2 phosphorylates the kinase Akt and dose-dependently activates Akt signaling (55).	GLP-2 stimulates nutrient absorption through the intracellular Akt-mTOR complex 1 signaling pathway (84).	The kinase Akt is known to regulate cell cycle, cell survival, cell growth, and cell metabolism, which are known hallmarks of cancer (48, 52, 61–63, 72, 73).
PI3K	Inhibition of PI3K blocks GLP-2–induced Akt phosphorylation (70).	PI3K inhibition reduces GLP-2–promoted nutrients uptake in the small intestine. Intestinal lysine uptake in the murine intestine was found to be dependent on PI3K/Akt-mTOR signaling (84).	GLP-2 dose-dependently phosphorylates Akt through the PI3K-dependent Akt-mTOR signaling pathway (70), which is a pathway associated with tumorigenesis (52, 73, 74).

Studies investigating GLP-2’s molecular mechanisms of action have mostly been conducted on rodents, pigs, and cell lines. The reviewed data suggest that GLP-2 most likely acts through second mediators such as amphiregulin, epiregulin, and heparin binding–epidermal growth factor, as well as induces Akt phosphorylation possibly dependent on PI3K signaling, which also a requires signaling pathway for the nutrients’ absorption in the murine intestine. IGF-I is considered the main mediator responsible for the GLP-2–induced intestinotrophic effects.

GLP-2 Induces Tumor Growth in Rodent Models

Different studies have aimed to clarify whether exogenous GLP-2 may have a growth-promoting effect on tumors in the gastrointestinal tract, especially after long-term administration.

To our knowledge, Thulesen et al. (80) was the first to investigate the effects of GLP-2 on colonic neoplasms in mice. Colonic adenomas were induced with 1,2-dimethylhydralazine (DMH) (80). The repeated subcutaneous DMH injections produced tubular adenomas predominantly in the lower part of the colon and rectum. After induction and a waiting period of 2 or 3 months, the mice were treated for 10 or 30 days with native GLP-2 or the DDP-4–resistant analogue Gly²–GLP-2. The data from the study showed that mice treated with Gly²–GLP-2 showed significantly increased growth and incidence of colonic neoplasms (80).

Another study by Iakoubov et al. (81) examined the tumor-promoting effects of exogenous GLP-2 but also endogenous GLP-2 using the GLP-2R antagonist GLP-2 (3–33). In the study, mutations in the murine intestine were initiated with azoxymethane. The study found significantly increased dysplastic changes in the group of mice treated with hGly²–GLP-2, suggesting a possible tumor-promoting effect of the GLP-2 analogue (81). Increased intestinal growth and proliferation were also observed among the hGly²–GLP-2–treated mice. However, these findings were only seen when the GLP-2 analogue was continuously administered, as the intestinotrophic effects disappeared 4 weeks after last treatment (81). Two of the GLP-2–treated mice developed tumors that the authors identified as adenocarcinomas with intact muscularis mucosa and no metastases (81). The tumors showed higher β-catenin expression and distribution than the adjacent normal colon tissue, as shown by IHC (81). Whether the findings of higher β-catenin expression and distribution in the tumors were mediated by GLP-2 treatment was not examined in the study.

Among the GLP-2 (3–33)–treated group, the incidence of dysplastic changes was decreased compared with that in the controls (81). The findings led the authors to conclude that the effects of exogenous GLP-2 are prominent at the early stage of carcinogenesis and not when adenomas or early dysplastic changes are already present in the intestine (81).

In a later study by Trivedi et al. (82), the exogenous and endogenous effects of GLP-2 were also investigated using the long-acting Gly²–GLP-2 and the antagonist GLP-2 (3–33). The rats were fed the carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine and a high-fat diet, whereas mice with dextran sodium sulfate–induced chronic colitis were injected with azoxymethane to promote two different models of inflammation-associated colon cancer (82). The study found a higher rate of dysplasia in both mice and rats treated with Gly²–GLP-2, whereas the antagonist GLP-2 (3–33) was found to reduce the incidence of high-grade dysplasia in the murine colon (82).

In the discussion of their study, Trivedi et al. (82) raised speculations that the overall effects of GLP-2 might be dependent on the type of rodent model used in GLP-2 tumor studies. This speculation was due to both their own findings and the results from a study by Koehler et al. (85).

The study by Koehler et al. (85) demonstrated no increase in adenoma formation or size in APC ^min/+ mice treated with native GLP-2 for 7 weeks. Based on these results, the study concluded that GLP-2 does not play a role in tumor growth or tumor formation (85). However, Koehler et al. (85) used only native GLP-2 to test the effects of the hormone.

A study from 2012 confirmed the findings of Thulesen et al. (80) from 2004. In that study, the aim was to examine the tumor-promoting effect of GLP-1 and sitagliptin in a mouse model of DMH-induced neoplasia. However, GLP-2 was included as a positive control. Only the mice treated with GLP-2 had a significantly increased number of adenomas (31).

In summary, a large proportion of the studies examining the neoplastic effects of the GLP-2 analogue in rodent models showed significant growth and tumor-promoting effects in the murine colon, especially when GLP-2 was administered for a longer time period (31, 80–82). However, the choice of model used seemed to affect the final results (30, 31, 80, 82).

Table 2 summarizes the reviewed studies on rodent models examining the effects of GLP-2 in relation to intestinal neoplasia.

Table 2.

Studies on Exogenously Administrated GLP-2’s Relation to Intestinal Growth and Intestinal Adenomas in Rodent Models

Study	Methods Used in Study	Results
Thulesen et al. (80)	Colonic adenomas were induced with DMH in mice. DMH injections produced tubular adenomas predominantly in the lower part of the colon and rectum.	The study found that mice treated with the exogenously administrated degradation-resistant GLP-2 analogue, Gly2–GLP-2, had significantly increased growth and incidence of colonic neoplasms.
Koehler et al. (85)	APC min/+ mice were treated with native GLP-2 for 7 weeks to investigate the tumor-promoting effects of exogenously administrated GLP-2.	The study found no alterations in adenoma formation or size and suggested that GLP-2 does not play a role in the tumor growth or tumor promotion in the murine intestine.
Iakoubov et al. (81)	Mutations in the murine intestine were initiated with AOM, and mice were later treated with either the antagonist GLP-2 (3–33) or the GLP-2 analogue, hGly2–GLP-2.	Results revealed significantly increased dysplastic changes among the mice treated with the GLP-2 analogue, hGly2-GLP. Increased intestinal growth and proliferation were also observed. The findings suggested a possible tumor-promoting effect of the GLP-2 analogue. The antagonist GLP-2 (3–33) was found to decrease the incidence of dysplasia in the murine intestine.
Trivedi et al. (82)	Rats were fed the carcinogen PhIP and a high-fat diet, whereas mice were fed with dextran sodium-sulfate. Both models were then injected with AOM.	A higher rate of dysplasia was noted in both models treated with the GLP-2 analogue, Gly2–GLP-2.
Kissow et al. 2012 (31)	The study examined the tumor-promoting effect of GLP-1 and sitagliptin in a mouse model of DMH-induced neoplasia. Exogenous GLP-2 administration was included as a positive control.	In the study, only the mice treated with the GLP-2 analogue had significantly increased incidence of intestinal adenomas.

Most of the studies examining the matter have found increased incidence of dysplastic changes in the murine intestine, suggesting that exogenous GLP-2 may play an important role in the development and progression of dysplastic changes as well as in the occurrence of adenomas in rodents treated with exogenous GLP-2.

Abbreviations: AOM, azoxymethane; PhIP, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine.

Therapeutic Benefits and Safety of Exogenous GLP-2 in Patients With SBS

The intestinotrophic effects of GLP-2 have also directed research toward the therapeutic potentials of the hormone in the treatment of gastrointestinal diseases, such as SBS, inflammatory bowel disease, and chemotherapy-induced mucositis. Several studies to date have identified various beneficial effects of GLP-2 in the intestines of both rodents and humans (5, 13, 14, 79, 86–91), leading to the approval of a degradation-resistant GLP-2 analogue (Gly²–GLP-2), teduglutide, for the treatment of SBS in 2012 (92).

One of the studies investigating the benefits of GLP-2–induced intestinal growth was a randomized placebo-controlled study consisting of 83 patients with SBS (5). The study examined the safety and tolerability of 24 weeks of teduglutide treatment and, to our knowledge, was characterized at that time as the first long-term randomized placebo-controlled study of teduglutide in patients with SBS dependent on parenteral nutrition (5). The study concluded that teduglutide treatment was well tolerated and improved the structure and function of the remaining intestine in patients with SBS as a result of the proabsorptive and intestinotrophic properties of GLP-2 (5). Colonoscopy was performed in 56 patients with colon in continuity. Of these patients, 45 were treated with teduglutide, and 11 were controls. No signs of malignancy or colonic adenomas were found in any of the groups (5). Following this research, 37 patients with colon in continuity proceeded in an extension study for a total of 52 weeks of treatment (93). However, this was an open-label, noncontrolled trial (93). Of the 37 patients, only a 22-year-old woman was found to have a 2-mm hyperplastic colonic adenoma and was withdrawn from the study. No cancers or adenomatous polyps were identified at the end of the study (93).

Between these studies, a randomized, placebo-controlled, multinational phase III study was conducted, showing a significantly reduced need for parenteral support among patients with SBS treated with 0.05 mg/kg teduglutide per day compared with placebo (12). The phase III study of 24 weeks concluded that teduglutide provided a new and effective treatment principle in the management of SBS; however, colonoscopy was only performed at baseline to exclude the presence of polyps or active intestinal diseases (12).

In 2016, the aforementioned phase III placebo-controlled trial was extended to a 2-year timespan, open-label, non-placebo-controlled study to examine the long-term effects of teduglutide treatment among patients with SBS (94). This study is currently the largest and longest study examining the tolerability and efficacy of teduglutide in humans (94).

Colonoscopies were performed within and at the end of the 24-month treatment period in patients with intact colons. At baseline colonoscopy, polyps were reported in nine patients, and all polyps were removed before treatment with teduglutide was initiated (94). Fifty patients received a study completion colonoscopy, and nine of these patients were found to have polyps (only one had a polyp removed at baseline). The patients with polyps were aged between 35 and 67 years, and 67% were female. Of these patients, seven had between one and five polyps in the colon/rectum (94). Two polyps were recorded as low-grade dysplasia (rectal/colorectal polyps), but none were recorded as overt malignancy (95).

One of the patients in the study had a history of Hodgkin lymphoma and developed a metastatic adenocarcinoma in the liver 11 months after starting teduglutide treatment (94). The study described the finding as not being linked to teduglutide treatment, as the origin of the metastasis was unknown, and the patient was predisposed to neoplasms due to earlier radiotherapy and chemotherapy treatments (94). However, the histopathology suggested that the metastasis most likely originated from a gastrointestinal tumor (94). The authors concluded that teduglutide is safe and tolerated but recommended that patients receiving teduglutide should regularly undergo colonoscopies for the early detection of potential gastrointestinal changes (94).

Because intestinal adenomas are known as potential precursors to cancer, the presence of gastrointestinal adenomas in teduglutide-treated patients raises a cause for concern. The team explained that the reported prevalence of polyps was within the recommended adenoma detection rate among first-time colonoscopies (94), but it could be discussed whether the results from colonoscopies performed at the end of the study should be considered first-time colonoscopies, as eight of the nine patients had no signs of adenomas at baseline. Because the long-term trial did not include a placebo arm, it could be reasonable to compare the results with data obtained by screening colonoscopies; however, data from screening programs are often biased by the inclusion criteria for screening (such as colonoscopy performed upon positive fecal blood screening). In addition, it must be emphasized that some patients included in the study were in fact high-risk patients, as the major intestinal resection could be secondary to both cancer and Crohn disease. No specific data on patients developing adenomas in the 2-year treatment period were reported.

Screening colonoscopies have been offered in Germany as a primary screening examination for the early detection and prevention of colorectal cancer since October 2002, and patients are referred for a subsequent colonoscopy 10 years later if the first was negative (96). The data extracted from the national registry reported that with a negative colonoscopy at the age of 55 years, there is a 20.4% (men) and 13.3% (woman) 10-year risk of developing any colorectal neoplasm (96).

We could not identify the data with a 2-year follow-up time after negative colonoscopy in patients with SBS or healthy subjects. Therefore, it is impossible to conclude if the reported polyps in the long-term study are simply patient related or are a consequence of teduglutide treatment.

Table 3 summarizes all the reviewed clinical studies on teduglutide treatment and whether the studies included colonoscopies among the teduglutide-treated patients.

Table 3.

Reviewed Studies on Teduglutide Treatment Among Patients With SBS and When Colonoscopies Were Performed on the Patients Undergoing Treatment

Study	Colonoscopies	Results
Jeppesen et al. (5)	Colonoscopies at baseline and after a 24-week treatment period	No adenomas detected
Jeppesen et al. (79)	Baseline colonoscopies only
Jeppesen et al. (12)	No colonoscopies performed
Tappenden et al. (91)	Baseline colonoscopies only
O’Keefe et al. (93)	Baseline colonoscopies as safety assessments and after 52 weeks of treatment	Of the 37 patients, only a 22-year-old woman was found to have a 2-mm hyperplastic colonic adenoma and was withdrawn from the study. No cancers or adenomatous polyps were identified at the end of the study.
Schwartz et al. (94)	Baseline colonoscopies and control colonoscopies in 50 patients after 24 months of teduglutide treatment	Colonic polyps were reported in 9 of 50 patients (18%). Two of these polyps were found before treatment started. Through histopathology, the polyps were identified as being benign, and no cases of dysplasia or malignancy were reported.
Rudzki et al. (95)	Baseline colonoscopies only

The findings of eventual colonoscopies among the studies are also summarized in the table. Most studies on teduglutide tolerability included a baseline colonoscopy, whereas few of the examined studies included a colonoscopy at the end of treatment. Adenomas found at baseline colonoscopies were removed before the start of teduglutide treatment, as the presence of adenomas was among the exclusion criteria for all the studies.

Concluding Remarks

Studies on rodents have shown that the administration of the GLP-2 analogue induces both intestinotrophic effects and neoplastic changes in the murine intestine (31, 80–82). No study longer than 30 months has been conducted in humans. Unfortunately, most of the studies do not include investigations of colonic neoplasms at baseline or at the end of teduglutide treatment.

At the molecular level, GLP-2 is found to induce its effects through secondary mediators, such as IGF-I and ErbB ligands, that are associated with cell growth (3, 48, 49, 60, 97, 98). Furthermore, studies have revealed that GLP-2 phosphorylates Akt, which is often dysregulated in many types of cancers, raising yet another cause for concern (52, 56, 61–64, 72, 73).

A recent study revealed that the observed beneficial effects of GLP-2 treatment among patients with SBS were associated with GLP-2–induced Akt–mTOR complex 1 activation (84). The study found that GLP-2 increased the absorption of amino acids in the murine intestine through the PI3K/Akt–mTOR complex 1 signaling pathway (84), suggesting an intracellular pathway responsible for the increased intestinal absorption observed among GLP-2–treated patients with SBS in clinical studies. The study also confirmed that GLP-2 activates the Akt-signaling pathway (84).

Taken together, both the cellular and molecular effects of GLP-2 are associated with hallmarks of tumorigenesis (77), explaining why the potential tumor-promoting effects of the GLP-2 analogue are of special interest in the research field.

The molecular effects of GLP-2 and its mechanisms of action in relation to tumorigenesis and SBS are summarized in Table 1.

Although 30 months of teduglutide treatment was concluded to be tolerable for patient with SBS, the connection between GLP-2 and intestinal tumor growth in humans remains unclear (5, 6, 93, 94, 99).

A clinical study that examined the efficacy and safety of 24 months of teduglutide treatment found adenomas in 9 of 50 patients during the final follow-up colonoscopy (94). It was uncertain whether the adenomas found among the nine patients were caused by long-term teduglutide treatment. The findings of this study raised yet another cause for concern, because intestinal adenomas are known as potential precursors to cancer (100–103). Whether longer periods of treatment with exogenous GLP-2 may play a role in the progression of intestinal cancers in patients with SBS has not been fully identified. The discussed clinical studies examining the long-term effects of teduglutide in the human intestine have all, however, recommended regular colonoscopy in patients who undergo teduglutide treatment (5, 12, 93, 94). Furthermore, in the prescribing information of teduglutide, the growth of neoplasms and the possibility of developing gastrointestinal polyps are listed as precautions (104), whereas immediate discontinuation of teduglutide is suggested when any gastrointestinal tumor is observed in patients undergoing teduglutide treatment (104).

It is conceivable that a change in GLP-2 levels and increased GLP-2R expression may play a role in the progression of intestinal cancers in humans, as the hormone is associated with mucosal growth and development of adenomas in the murine intestine (31, 80–82). As exogenous GLP-2 is associated with growth and antiapoptosis on a molecular level and preneoplastic changes on a macroscopic level, the hormone’s use in therapeutic purposes should be carried out with caution. Special attention must be given to patients who are already predisposed to gastrointestinal cancers.

In this narrative review, we have presented and discussed existing studies concerning the intestinotrophic effects of GLP-2 and the hormone’s relation to intestinal neoplasia. Because most of the molecular effects of GLP-2 are not fully understood, future research examining these matters is needed. In particular, long-term clinical studies on humans are required, as most studies to date have used either rodents or cell lines as models.

The uncertainty around GLP-2’s relation to intestinal neoplasia in humans will continue to raise concerns regarding the safety of long-term and lifelong treatments with GLP-2 analogues until further research is conducted.

Abbreviations:

Abbreviations:
 
• DMH

  1,2-dimethylhydralazine

 
• DPP-4

  dipeptidyl peptidase-4

 
• EGFR

  epidermal growth factor receptor

 
• GLP-1

  glucagon-like peptide-1

 
• GLP-2

  glucagon-like peptide-2

 
• GLP-2R

  glucagon-like peptide-2 receptor

 
• GSK-3β

  glycogen synthase kinase-3β

 
• IHC

  immunohistochemistry

 
• mRNA

  messenger RNA

 
• PI3K

  phosphatidylinositol 3-kinase

 
• SBS

  short bowel syndrome

Acknowledgments

Valuable and constructive suggestions given by clinical lecturer Lasse Bremholm Hansen were deeply appreciated.

Disclosure Summary: The authors have nothing to disclose.