sign in sign up
<<
Home , Incretin, Metformin, Sitagliptin, Thiazolidinedione, Vildagliptin

Bench to Clinic Symposia EDITORIAL REVIEW

Dipeptidyl Peptidase-4 Inhibitors Clinical data and clinical implications

BO AHREN´ , MD, PHD TWO STRATEGIES FOR GLP-1– BASED THERAPY — To harness the antidiabetic action of GLP-1 and at the he dipeptidyl peptidase-4 tion and stimulates secretion (1). same time overcome the problem of the (DPP-4) prevents the inactivation of GLP-1 also exhibits strong antidiabetic rapid inactivation of the native hor- T -like -1 (GLP-1). actions, as initially demonstrated already mone, two strategies have been ex- Since GLP-1–based therapy is a promis- in the early 1990s (2–4). Thus, infusion plored. Both these strategies have been ing novel treatment of type 2 , the of GLP-1 lowers circulating shown to be successful. One approach strategy to inhibit the enzyme has been through a combination of stimulation of is to use GLP-1 receptor agonists explored. Several DPP-4 inhibitors are in insulin secretion and inhibition of gluca- (GLP-1 mimetics), which are not de- clinical development; these are orally ac- gon secretion. A 6-week study with con- graded by DPP-4. A representative of tive and increase levels of active GLP-1, tinuous subcutaneous infusion of GLP-1 this approach is (Byetta; Lilly) which in turn increases insulin secretion showed reduction in and prandial (8), which is now approved for use in and reduces glucagon secretion and glycemia along with reduction in A1C the treatment of diabetes both in the U.S. and Europe. (Novo thereby lowers glucose levels. Most expe- and improvement both in insulin secre- Nordisk) is another example of this rience exists for (Merck) and tion and insulin action and reduction in strategy. The other approach for GLP- (), which both have a body weight, which illustrates the po- 1–based therapy is to inhibit the en- long duration of action, allowing once- tency of GLP-1–based therapy (5). daily administration. In drug-naı¨ve sub- zyme activity of DPP-4 (1,9–13). jects with , both sitagliptin and vildagliptin reduce A1C levels by DPP-4 INACTIVATION OF ϳ1% as monotherapy, as demonstrated DPP-4 inhibition as a strategy to GLP-1 — It was early understood that treat diabetes in studies up to 52 weeks. Also in combi- GLP-1 is unattractive as chronic therapy nation with and thiazo- The rationale for the strategy of inhibiting of diabetes because the is rap- DPP-4 in the treatment of type 2 diabetes lidinediones, sitagliptin and vildagliptin idly inactivated by the action of the en- improve glycemic control with reduction is to prevent the inactivation of GLP-1 and ϳ zyme DPP-4 (6). This enzyme is widely therefore to enhance and prolong the ac- of A1C of 1%. Both sitagliptin and expressed in several organs and circulates vildagliptin are safe and tolerable with tion of the endogenously released in a soluble form (7). It acts by cleavage of hormone. This strategy was first summa- low risk of . They are both the two NH -terminal amino acids of bio- body weight neutral. The studies pre- 2 rized by Holst and Deacon (14), who active , provided that the second showed that DPP-4 inhibition increases sented thus far therefore suggest that amino acid is alanine or proline. Since the DPP-4 inhibition is an efficient treatment circulating levels of GLP-1 in experimen- second NH2-terminal amino acid in tal animals and that the insulinotropic ac- of type 2 diabetes, both as monotherapy GLP-1 is alanine, GLP-1 is cleaved to a and combination therapy. Because of its tion of exogenously administered GLP-1 truncated form [sometimes called GLP-1 efficiency, safety, and tolerability in associ- is augmented by DPP-4 inhibition. Later (9-36)amide] (6). This truncated form of ation with the oral mode of administration, studies demonstrated that the prevention GLP-1 is largely inactive; therefore, the it is expected that DPP-4 inhibition will be a of inactivation of GLP-1 by DPP-4 inhibi- cleavage of GLP-1 by DPP-4 is an inacti- first-line treatment of the early stage of type tion markedly increases the active GLP-1 vation process. The cleavage is rapid, 2 diabetes, particularly in combination with in the circulation. For example, as shown metformin or . which is the reason why native GLP-1 has in dogs, the active GLP-1 under fasting a short half-life (Ͻ2 min). Therefore, it is conditions accounts for only ϳ10% of to- GLP-1 AS AN necessary to use multiple daily injections tal GLP-1, whereas after DPP-4 inhibition ANTIDIABETIC HORMONE — of continuous subcutaneous infusion of 99% of GLP-1 is in the active form (15). GLP-1 is one of the important incretin GLP-1 for maintenance of glycemic con- The increase in concentrations of active ; it is released after meal inges- trol when using the native GLP-1. GLP-1 following administration of DPP-4 inhibitors has subsequently been con- ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● firmed in clinical studies (Fig. 1). It has From the Department of Clinical Sciences, Division of Medicine, Lund University, Lund, Sweden. also been demonstrated that DPP-4 inhi- Address correspondence and reprint requests to Dr. Bo Ahre´n, MD, PhD, Lund University, Division of bition increases not only prandial but also Medicine, B11 BMC, SE-221 84 Lund, Sweden. E-mail: [email protected]. Received for publication 4 February 2007 and accepted in revised form 11 February 2007. fasting levels of active GLP-1 (16). In fact, Published ahead of print at http://care.diabetesjournals.org on 2 March 2007. DOI: 10.2337/dc07-0233. DPP-4 inhibition results in an overall in- B.A. has been a consultant for Novartis and Merck. crease in GLP-1 levels with preserved cir- Abbreviations: DPP-4, dipeptidyl peptidase-4; FDA, Food and Drug Administration; GIP, glucose- cadian rhythm throughout the day (Fig. dependent insulinotropic polypeptide; GLP-1, glucagon-like peptide-1. A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion 2). Since DPP-4 is also the inactivation factors for many substances. enzyme for the other incretin hormone, © 2007 by the American Diabetes Association. glucose-dependent insulinotropic

1344 DIABETES CARE, VOLUME 30, NUMBER 6, JUNE 2007 Ahre´n

It was found that following a 4-week treatment period, fasting, prandial, and mean glucose levels were reduced by Ͼ1 mmol/l, and even though the study was only 4 weeks in duration, A1C was low- ered by 0.5%. The treatment was also highly tolerable and safe, suggesting the feasibility of this approach for the treat- ment of type 2 diabetes.

Vildagliptin and sitagliptin Following these initial animal studies and the successful proof-of-concept study in humans, several DPP-4 inhibitors have been developed, which are in different stages in clinical development (Table 1). Vildagliptin (LAF237, Galvus; Novartis) and sitagliptin (MK-0431, Januvia; Merck) have been explored in detail (23– 25). They are orally active and rapidly ab- sorbed, and both efficiently inhibit plasma DPP-4 activity—plasma DPP-4 activity is inhibited by almost 100% al- ready at 15–30 min after oral administra- tion, and Ͼ80% inhibition lasts for Ͼ16 h (26). They are therefore both possible to administer once daily; sitagliptin has been approved by the Food and Drug Admin- istration (FDA) (October 2006), whereas vildagliptin has been applied for ap- proval. Both compounds are planned to be used in a dose of 100 mg once daily. Whereas hepatic insufficiency does not seem to alter of the Figure 1—Plasma levels of active GLP-1, glucose, insulin, and glucagon before and after intake compounds, renal insufficiency increases of a standardized breakfast (at time 0) after 4 weeks of treatment with vildagliptin (100 mg daily; circulating sitagliptin (28). Therefore, in n ϭ 18) or placebo (n ϭ 19) in subjects with type 2 diabetes (reproduced with permission from the patients with moderate (creatinine clear- Endocrine Society, ref. 26). ance Ͻ50 ml/min) or severe (creatinine Ͻ30 ml/min) renal insuffi- polypeptide (GIP) (5), the concentrations mean A1C 7.4%, and mean fasting glu- ciency, the dose of sitagliptin should be of active GIP are also increased through- cose 9.0 mmol/l) were treated with the reduced to 50 and 25 mg, respectively, out the 24-h period after DPP-4 inhibition DPP-4 inhibitor NVP-DPP728 (Novartis). and the FDA recommends that renal func- (16). This is, however, probably of less importance for the antidiabetic action of DPP-4 inhibition, since GIP seems to have lost much of its insulinotropic action in diabetes and, furthermore, since GIP stimulates rather than inhibits glucagon secretion (17).

Early studies on DPP-4 inhibition The strategy to inhibit DPP-4 as a treat- ment of diabetes was initially verified in animal studies, which demonstrate that genetic deletion of DPP-4 (18) or pharma- cological inhibition of DPP-4 improves glucose tolerance and insulin secretion in a variety of experimental models (12,19– 21). The final proof-of-concept study in Figure 2—Plasma levels of intact GLP-1 during 13.5-h sampling comprising three standardized humans was published in 2002 (22). Pa- meals the day before and after 28 days’ treatment with vildagliptin (100 mg twice daily) in nine tients with drug-naı¨ve type 2 diabetes drug-naı¨ve patients with type 2 diabetes (reproduced with permission from the Endocrine Society, (mean age 65 years, mean BMI 27 kg/m2, ref. 16).

DIABETES CARE, VOLUME 30, NUMBER 6, JUNE 2007 1345 DPP-4 inhibitors

Table 1—DPP-4 inhibitors in various stages of clinical development according to various pared with treatment with metformin at databases in the public domain 1 g twice daily (n ϭ 254) in subjects with type 2 diabetes (33). The subjects were, Stage in again, drug-naı¨ve patients with a mean Name Company development diabetes duration of 2.4 years, and they had a mean baseline A1C of 8.7%. It was Sitagliptin (Januvia) Merck Approved by FDA found that vildagliptin reduced A1C by Vildagliptin (Galvus) Novartis Filed to FDA 1.0%, whereas metformin reduced A1C Takeda Phase III by 1.4%; hence, although noninferiority Bristol-Myers Squibb Phase III versus metformin was not reached, vilda- PSN-9301 OSI Pharmaceuticals Phase II gliptin also showed a clinically meaning- R1438 Roche Phase II ful reduction of A1C in this study. Again, TA-6666 Tanabe Phase II vildagliptin was safe and tolerable; three PHX1149 Phenomix Phase II patients reported mild hypoglycemic GRC 8200 Glenmark Pharmaceuticals Phase II events in the vildagliptin plus metformin– SYR-619 Takeda Phase I treated group versus one patient in the TS-021 Taisho Pharmaceuticals Phase I metformin-treated group. SSR 162369 Sanofi-Aventis Phase I Sitagliptin. The first study using sita- ALS 2-0426 Alantos Pharmaceuticals Phase I gliptin as monotherapy evaluated the compound in 743 drug-naı¨ve patients with type 2 diabetes using different doses tion is assessed before start of sitagliptin pendency of the action. Another study for a treatment duration of 12 weeks (34). treatment. No drug interactions have showed that during a 24-week period in The patients had a mean baseline A1C of been observed for the DPP-4 inhibitors 354 drug-naı¨ve patients with type 2 dia- 7.7% and were inadequately treated with and other drugs. betes (baseline A1C 8.4%), A1C was re- diet and exercise. Sitagliptin was given at duced by 0.5% by vildagliptin at 50 mg the doses of 5, 12.5, 25, or 50 mg twice Vildagliptin and sitagliptin as once daily, by 0.7% by vildagliptin at 50 daily. It was found that after 12 weeks, all monotherapy mg twice daily, and by 0.9% by vildaglip- groups given sitagliptin had displayed a Vildagliptin. In the first clinical study tin at 100 mg once daily (31). Vildagliptin significant reduction in A1C in relation to with vildagliptin, 100 mg once daily was was well tolerated in all groups. This placebo, with the largest reduction versus administered to 18 drug-naïve patients shows that 100 mg vildagliptin provides placebo (0.77%) at the dose of 50 mg with type 2 diabetes for 4 weeks and the similar clinical benefit when given as sin- twice daily. In one arm of the study, glipi- effects compared with a placebo- gle dose or divided. zide was given at 5 mg daily with elective administered group (n ϭ 19). The pa- In two studies, vildagliptin as mono- titration to 20 mg daily. In this group, tients had a mean baseline glucose of 8.8 therapy has been compared with other ac- A1C was reduced by 1.0%. Hypoglyce- mmol/l and a mean baseline A1C of 7.2% tive treatment modalities. A 24-week mic events were observed in 17% of pa- (26). It was found that vildagliptin re- study compared the effect of vildagliptin tients given versus Ͻ1% in the duced fasting glucose by ϳ0.7–0.9 (50 mg twice daily; n ϭ 519) with that of sitagliptin groups. Whereas body weight mmol/l and prandial glucose by 1.5 (8 mg once daily; n ϭ 267) increased in subjects given glipizide (by mmol/l, and A1C also was significantly in subjects with type 2 diabetes (32). The 1.1 kg), no significant change in body reduced (26). A subsequent study in pa- subjects were drug-naı¨ve patients with a weight was observed in subjects given tients with type 2 diabetes with a mean mean baseline A1C level of 8.7% and a sitagliptin. Another study examined the baseline A1C of 7.7% showed that after mean diabetes duration of 2.4 years. Fol- influence of sitagliptin as monotherapy 12-week treatment with vildagliptin at 50 lowing the 24-week treatment, vildaglip- over 12 weeks versus placebo; a total of or 100 mg once daily, A1C had been re- tin reduced A1C by 1.1%. The reduction 552 patients were included. They were all duced by 0.46 and 0.40%, respectively, was seen during the first 5–10 weeks, and drug-naı¨ve patients and had a mean base- compared with the nonsignificant 0.13% then a sustained effect of vildagliptin was line A1C level of 7.7%. Sitagliptin was reduction of A1C in the placebo group evident. The reduction in A1C was not given at 25, 50, or 100 mg once daily or (29). In a third study, vildagliptin was significantly different from the reduction 50 mg twice daily, and it was found that given as monotherapy at 25 mg twice with rosiglitazone, which was 1.3%. the largest reduction in A1C (by 0.6%) daily for 12 weeks (30). It was demon- Vildagliptin was safe and tolerable and was observed by the dose of 100 mg daily strated that both fasting and prandial glu- showed a lower degree of adverse events (35). This was accompanied by a reduc- cose were reduced and that A1C was than rosiglitazone. One mild hypoglyce- tion in fasting glucose by 0.9 mmol/l. Fur- reduced by 0.6% from a mean baseline of mic event was evident in each group, and thermore, an 18-week study comprising 8.0%. These placebo-controlled studies whereas patients given rosiglitazone ex- 521 patients using sitagliptin at 100 or thus showed that vildagliptin efficiently perienced an increase in body weight by 200 mg once daily as monotherapy versus improved glycemic control. Moreover, 1.6 kg, those given vildagliptin had no placebo in patients with an initial A1C vildagliptin was safe and tolerable with significant change in body weight. The in- level of 8.1% showed a reduction in A1C adverse events of the same degree as in the cidence of edema was greater with rosiglit- by 0.60 and 0.48%, respectively (36; Fig. placebo groups. Furthermore, the num- azone (4.1%) than vildagliptin (2.1%). 3). Finally, a 24-week study using sita- ber of hypoglycemic events was very low, Vildagliptin as monotherapy at 50 mg gliptin at 100 or 200 mg daily in drug- which is expected from the glucose de- twice daily (n ϭ 526) is also being com- naı¨ve patients with type 2 diabetes with a

1346 DIABETES CARE, VOLUME 30, NUMBER 6, JUNE 2007 Ahre´n

metformin when compared with the con- trol group given metformin alone; the re- duction in A1C was 0.7% in the group given 50 mg once daily and 1.1% in the group given vildagliptin at 50 mg twice daily. Vildagliptin has also been examined when used in combination with pioglita- zone in a 6-month study involving 592 patients who were assigned to treatment with vildagliptin alone at 100 mg daily, with vildagliptin at 50 or 100 mg in com- bination with at 15 or 30 mg, respectively, or with pioglitazone alone at 30 mg daily (41). Baseline A1C was 8.7%, and a large reduction in A1C (1.9%) was Figure 3—A1C over time in patients with type 2 diabetes given placebo (E), once-daily sitagliptin observed in the group treated with vilda- at 100 mg (), and once-daily sitagliptin at 200 mg (f) (reproduced with permission from gliptin at 100 mg in combination with Springer Verlag, ref. 36). pioglitazone at 30 mg daily. Finally, vilda- gliptin has also been added to insulin in subjects with more advanced type 2 dia- mean baseline A1C of 8.0% showed a re- A1C levels by 0.7% compared with met- betes in a 24-week study (42). In this duction by 0.74 and 0.94%, respectively formin alone. Furthermore, during the study, mean duration of diabetes was (37). In conclusion, studies using vilda- following 40 weeks, A1C increased by 14.6 years and patients had been treated gliptin or sitagliptin as monotherapy in 0.066% per month in patients given met- with insulin for a mean duration of 6.3 patients with type 2 diabetes show good formin alone, whereas in subjects given years. The mean daily insulin dose was 82 efficiency to improve the glycemic control vildagliptin in combination with met- units, and mean baseline A1C was 8.9%; with a reduction in A1C of ϳ0.8–1.1% formin the rate of increase in A1C was this was reduced by 0.5% in the group over 12–52 weeks of treatment. As in only 0.013% per month (Fig. 4). This given vildagliptin with insulin versus most studies in type 2 diabetes, subjects suggests the progressive deterioration in 0.2% in the group given insulin alone. with the highest baseline A1C levels have glycemic control seen in patients treated Hypoglycemic events were less common the largest reduction in A1C; therefore, with metformin was prevented by addi- and less severe in the patients given vilda- when directly comparing the degree of re- tion of vildagliptin. A larger study com- gliptin in combination with insulin (33 duction, this must be taken into account. prising altogether 416 patients examined patients, 113 events, 0 severe events) ver- Finally, in monotherapy there is no indi- the addition of vildagliptin at 50 mg once sus in those given insulin alone (45 pa- cation of differences in efficacy between or twice daily to ongoing treatment with tients, 185 events, 6 severe events). younger and older subjects and no rela- metformin (40). The patients had a mean Hence, in combination with insulin, tion to degree of , although results diabetes duration of 6.2 years and a mean vildagliptin reduced A1C more markedly in these subsets of patients are not always baseline A1C of 8.4%. The results show, than insulin alone, and this was associ- clearly reported. again, a reduction in A1C in the groups ated with a reduced risk of hypoglycemia. given vildagliptin in combination with Also, sitagliptin has been evaluated in Vildagliptin and sitagliptin in combination therapy Vildagliptin. Studies have reported the experience of treatment with a DPP-4 in- hibitor in combination with metformin, a , or insulin. The first study with vildagliptin in combination therapy was a 52-week trial in which the compound was added at the dose of 50 mg daily to ongoing treatment with met- formin (39). The patients had a mean age of 57 years, a mean diabetes duration of 5.5 years, a mean of 2.3 years of ongoing treatment with metformin (1.5–3.0 g daily), and a mean baseline A1C of 7.8%. The study was designed as an initial core study for 12 weeks (n ϭ 107), which was followed by a 40-week extension period in a subset of patients (n ϭ 71). The re- sults showed that during the initial 12- Figure 4—Time course of A1C over 52 weeks in subjects with type 2 diabetes treated with week study period, vildagliptin in vildagliptin (50 mg once daily) in combination with metformin (n ϭ 42) versus metformin alone combination with metformin reduced (n ϭ 29) (reproduced with permission from ref. 38).

DIABETES CARE, VOLUME 30, NUMBER 6, JUNE 2007 1347 DPP-4 inhibitors combination with metformin. In a 6-month tion of 2 vs. Ϫ0.3 mmHg in the placebo- of DPP-4 study comprising 701 patients, sitagliptin administered group; the difference being inhibition (100 mg daily) was added to ongoing met- significant) (40). This deserves further DPP-4 inhibition prevents the inactiva- formin (Ͼ1.5 g daily) in subjects with a study. Furthermore, frequency of electro- tion of GLP-1, and this increases GLP-1 mean baseline A1C of 8.0%. A1C was re- cardiogram abnormalities during treatment levels. The increase in GLP-1 levels is seen duced by 0.65% by sitagliptin compared with DPP-4 inhibitors has not been higher throughout 24 h, i.e., both after meal in- with placebo (43). In another large study than in the placebo-administered groups. gestion and in the fasting state (16). The comprising a total of 1,172 patients, the ef- Moreover, in contrast to the reduction in importance of GLP-1 for the action of fect of sitagliptin (100 mg daily) in combi- body weight seen after treatment with DPP-4 inhibitors is evident from animal nation with metformin was compared with GLP-1 analogs, DPP-4 inhibitors are body studies showing that DPP-4 inhibition glipizide (up to 20 mg daily) as add-on to weight neutral. This body weight neutrality does not improve glucose homeostasis in metformin during a study period of 52 distinguishes this class of compound from mice with genetic deletion of GLP-1 and weeks. The included subjects had a mean the increase in body weight, which is asso- GIP receptors (47). GLP-1 in turn stimu- baseline A1C of 7.5%, and it was reported lates insulin secretion, and it has been ciated with treatment with thiazolidinedi- that in both groups A1C was reduced by documented that acute ␤-cell function is ones and . 0.67% (44). The study reported two impor- improved also by DPP-4 inhibition. This A potential concern during the devel- tant differences between the groups—the has been documented for vildagliptin as number of hypoglycemic events was strik- opment of DPP-4 inhibition as a novel increased insulin response in relation to ingly higher in the group given glipizide therapy has been the potential of DPP-4 to the glucose response after meal ingestion (32%) than in the group given sitagliptin also cleave other bioactive peptides with (26) and increased estimated insulin se- (4.9%), and body weight increased in sub- alanine or proline as the second amino cretory rate after modeling insulin and C- jects given glipizide (by 1.5 kg) but de- acid from the NH2-terminal end (7). Be- peptide data after meal ingestion (16,48). creased in subjects given sitagliptin (by 1.5 cause these other bioactive peptides may Furthermore, sitagliptin has been shown kg). Sitagliptin (100 mg daily, n ϭ 175) has include , -releasing to increase homeostasis model assess- also been examined versus placebo (n ϭ peptide, substance P, and various chemo- ment-B index (a marker for insulin secre- 178) as add-on to pioglitazone at 30 or 45 kines, this phenomenon may potentially tion) and to reduce proinsulin-to-insulin mg daily for 6 months in patients with a cause adverse events related to increased ratio (a marker for ␤-cell function) (36). diabetes duration of 6.1 years and a mean , neurogenic inflamma- Animal studies have also shown im- baseline A1C of 8.0% (45). Compared with tion, and immunological reactions. How- proved chronic ␤-cell function, such as placebo, A1C was reduced by 0.7% by sita- ever, no such adverse events have been increased ␤-cell mass, after DPP-4 inhibi- gliptin in combination with pioglitazone. reported in animal studies or in humans tion (12,19,21). However, no such evi- Sitagliptin was well tolerated, and the pro- using DPP-4 inhibition. It should also be dence exists in humans. insulin-to-insulin ratio was reduced in the emphasized that it has not been demon- Another important mechanism for group treated with sitagliptin. strated that DPP-4 indeed affects the me- improved glycemic control by DPP-4 in- The studies presented thus far with tabolism of these other bioactive peptides hibition is inhibition of glucagon secre- vildagliptin and sitagliptin in combina- in humans; studies referred to previously tion, which is an effect by GLP-1 as well. tion therapy therefore show good efficacy are mainly in vitro studies in test tubes. This was initially demonstrated when in combination with both metformin and vildagliptin at 100 mg once or twice daily thiazolidinediones in studies of at least 6 was given for 4 weeks (26), and it has also months’ duration. Mean A1C levels are DPP-4 inhibitors and levels been shown that the entire 24-h glucagon ϳ reduced by 0.6–1%, with the highest The clinical trials with vildagliptin and profile is reduced by vildagliptin (16). reduction in subjects with highest base- sitagliptin have measured fasting levels of Furthermore, a recent study showed that line A1C levels. , and in general no or very little ef- the reduction in glucagon levels in asso- fects have been found on parameters such ciation with increased insulin secretion by Safety and tolerability of DPP-4 as cholesterol and . This administration of 100 mg vildagliptin to inhibitors subjects with type 2 diabetes was accom- would suggest that DPP-4 inhibition does Both vildagliptin and sitagliptin are tolera- panied by inhibition of hepatic glucose not affect lipid . However, a ble and safe with an adverse events profile production, as determined by clamp tech- recent study (46) showed that vildagliptin similar to that of placebo-administered pa- nique using tracer glucose (49). These tients in clinical studies. Also, the reported largely (by 85%) inhibited the findings are of great importance consider- numbers of hypoglycemia are very low dur- response to ingestion of a fat-rich meal ing the inappropriately high glucagon se- ing DPP-4 inhibition, which is expected following a 4-week treatment at 50 mg cretion in subjects with type 2 diabetes from the glucose dependency on the effects twice daily in patients with type 2 diabe- (50). of GLP-1. Reports on effects of DPP-4 inhi- tes. This was seen in association with a Besides the improvement of islet bition on blood pressure are few, and when 91% reduction in chylomicron apoli- function, DPP-4 inhibition may also im- reported, no 24-h measurements have been poprotein B-48 and chylomicron choles- prove insulin sensitivity. This has been undertaken. Most studies report no effect terol. Hence, vildagliptin improves found following treatment with vilda- on blood pressure. However, in one 24- plasma lipids and lipoprotein particle me- gliptin using both an indirect measure week study with vildagliptin in combina- tabolism after a fat-rich meal, which of insulin sensitivity (48) and the hyper- tion with metformin, a slight reduction in would be of importance considering the insulinemic-euglycemic clamp test blood pressure was seen in the group given relevance for prandial lipemia as a marker (51). This may be a consequence of the vildagliptin (systolic blood pressure reduc- for cardiovascular diseases. reduced glucagon levels seen after

1348 DIABETES CARE, VOLUME 30, NUMBER 6, JUNE 2007 Ahre´n

DPP-4 inhibition in association with the long-term durability and long-term tin hormones by inhibition of dipeptidyl improved insulin action due to im- safety as well as action in subsets of pa- peptidase IV suppresses meal-induced in- proved metabolic control. tients, such as in the elderly. cretin secretion in dogs. J Endocrinol 172: In contrast to GLP-1, DPP-4 inhibi- 355–362, 2002 tion does not seem to affect gastric emp- 16. Mari A, Sallas WM, He YL, Watson C, References Ligueros-Seylan M, Dunning BE, Deacon tying, as is evident by the lack of effect of CF, Holst JJ, Foley JE: Vildagliptin, a DPP-4 inhibitors on the rate of increase in 1. Drucker DJ, Nauck MA: The incretin sys- tem: glucagon-like peptide-1 receptor dipeptidyl peptidase-IV inhibitor, im- circulating glucose after meal ingestion. agonists and dipeptidyl peptidase-4 in- proves model-assessed beta-cell function Furthermore, vildagliptin did not affect hibitors in type 2 diabetes. Lancet 368: in patients with type 2 diabetes. J Clin En- the rate of gastric emptying of a tracer- 1696–1705, 2006 docrinol Metab 90:4888–4894, 2005 enriched meal (52). 2. Gutniak M, Ørskov C, Holst JJ, Ahre´n B, 17. Meier JJ, Nauck MA: GIP as a potential Efendic S: Antidiabetic effect of glucagon- therapeutic agent? Horm Metab Res 36: SUMMARY — Several DPP-4 inhib- like peptide-1 (7–36) amide in normal 859–866, 2004 itors are in clinical development; one subjects and patients with diabetes melli- 18. Marguet D, Baggio L, Kobayashi T, Ber- (sitagliptin) has been approved by the tus. N Engl J Med 326:1316–1322, 1992 nard AM, Pierres M, Nielsen PF, Ribel U, 3. Nathan DM, Schreiber E, Fogel H, Mojsov Watanabe T, Drucker DJ, Wagtmann: En- FDA, and another (vildagliptin) has been hanced insulin secretion and improved applied for approval. Both of these com- S, Habener JF: Insulinotropic action of glucagonlike peptide-1-(7-37) in diabetic glucose tolerance in mice lacking CD26. pounds are efficient with a remarkably and nondiabetic subjects. Diabetes Care Proc Natl Acad SciUSA97:6874–6879, low degree of adverse events. Although no 15:270–276, 1992 2000 head-to-head studies have been reported, 4. Nauck MA, Wollschlaeger D, Werner J, 19. Burkey BE, Li X, Bolognese KL, Balkan B, the reported studies suggest that these Holst JJ, Ørskov C, Creutzfeldt W, Willms Mone M, Russell M, Hughes TE, Wang two compounds have similar effects on B: Effects of subcutaneous glucagon-like PR: Acute and chronic effects of the incre- glycemic control. Therefore, differences peptide 1 (GLP-[7–36amide]) in patients tin enhancer vildagliptin in insulin-resis- between studies relate more likely to the with NIDDM. Diabetologia 39:1546– tant rats. J Pharmacol Exptl Ther 315:688– patients in the trials than to the inherent 1553, 1996 695, 2005 differences between the compounds. The 5. Zander M, Madsbad S, Madsen JL, Holst 20. So¨rhede Winzell M, Ahre´n B: The high-fat fed mouse: a model for studying mecha- overall experience with DPP-4 inhibition JJ: Effect of 6-week course of glucagon- like peptide 1 on glycaemic control, insu- nisms and treatment of impaired glucose is therefore that they are orally active, lin sensitivity, and ␤-cell function in type tolerance and type 2 diabetes. Diabetes 53 safe, and highly tolerable, with a minimal 2 diabetes: a parallel-group study. Lancet (Suppl. 3):S215–S219, 2003 risk for hypoglycemic events. Further- 359:824–830, 2002 21. Mu J, Woods J, Zhou YP, Roy RS, Li Z, more, they show sustained, robust, and 6. Deacon CF: Circulation and degradation Zycband E, Feng Y, Zhu L, Li C, Howard clinically significant improvement in of GIP and GLP-1. Horm Metab Res 36: AD, Moller DE, Thornberry NA, Zhang glycemia in both monotherapy and 761–765, 2004 BB: Chronic inhibition of dipeptidyl pep- combination with metformin and thia- 7. Mentlein R: Dipeptidyl-peptidase IV tidase-4 with a sitagliptin analog pre- ␤ zolidinediones, and they are body (CD26): role in the inactivation of regula- serves pancreatic -cell mass and weight neutral. The relevant mechanisms tory peptides. Regul Pept 85:9–24, 1999 function in a rodent model of type 2 dia- betes. Diabetes 55:1695–1704, 2006 of action of DPP-4 inhibition (mainly im- 8. Ahre´n B: Exenatide: a novel treatment of type 2 diabetes. Therapy 2:207–222, 2005 22. Ahre´n B, Simonsson E, Larsson H, provement of islet function) and the effi- 9. Ahre´n B: Inhibition of dipeptidyl pepti- Landin-Olsson M, Torgeirsson H, Jansson cacy, tolerability, and safety of treatment dase-4 (DPP-4): a novel approach to treat PA, Sandqvist M, Båvenholm P, Efendic S, suggest that this approach has great po- type 2 diabetes. Curr Enz Inhib 1:65–73, Eriksson JW, Dickinson S, Holmes D: In- tential as a novel treatment. Its place in 2005 hibition of dipeptidyl peptidase IV im- therapy remains, however, to be explored 10. Barnett A: DPP-4 inhibitors and their po- proves metabolic control over a 4-week in more detail. Results presented thus far tential role in the management of type 2 study period in type 2 diabetes. Diabetes suggest that DPP-4 inhibition has its diabetes. Int J Clin Pract 60:1454–1470, Care 25:869–875, 2002 strongest potential as a first-line treatment 2006 23. Villhauer EB, Brinkman JA, Naderi GB, Bur- in early stages of type 2 diabetes in com- 11. Deacon CF, Ahre´n B, Holst JJ: Inhibitors key BF, Dunning BE, Prasad K, Mangold BL, Russell ME, Hughes TE: 1[[(3-hydroxy-1- bination with metformin and thiazo- of dipeptidyl peptidase IV: a novel ap- proach to prevention and treatment of adamantyl)amino-]acetyl]-2-cyano-(s)-pyr- lidinediones. DPP-4 inhibitors are also type 2 diabetes. Expert Opin Investig Drugs rolidine: a potent, selective, and orally strong candidates for being established as 13:1091–1102, 2004 bioavailable dipeptidyl peptidase IV inhibi- a first-line treatment as monotherapy, 12. Drucker DJ: Biologic actions and thera- tor with antihyperglycemic properties. particularly in elderly subjects and in sub- peutic potential of the - J Med Chem 46:2774–2789, 2003 jects with contraindication or intolerance derived peptides. Nature Clin Pract 24. Kim D, Wang L, Beconi M, Eiermann GJ, for metformin or thiazolidinediones. Endocrinol Metab 1:22–31, 2005 Fisher MH, He H, Hickey GJ, Kowalchick DPP-4 inhibition may also be used in 13. Ahre´n B, Schmitz O: GLP-1 receptor ago- JE, Leiting B, Lyons K, Marsillo F, Mc- combination with insulin with the great nists and DPP-4 inhibitors in the treat- Cann ME, Patel RA, Petrov A, Scapin G, advantage of reducing the likelihood of ment of type 2 diabetes. Horm Metab Res Patel SB, Roy RS, Wu JK, Wyvratt MJ, developing hypoglycemia. Whether 36:867–887, 2004 Zhang BB, Zhu L, Thornberry NA, Weber 14. Holst JJ, Deacon CF: Inhibition of the ac- AE: (2R)-4-oxo-4-[3-(trifluoromethyl)- DPP-4 inhibition may also be advanta- tivity of dipeptidyl-peptidase IV as a treat- 5,6-dihydro[1,2,4]triazol[4,3,a]pyrazin- geous over existing treatment in long- ment for type 2 diabetes. Diabetes 47: 7(8H)-yl]-1-(2,4,5- term therapy of more advanced stages of 1663–1670, 1998 trifluorophenyl)butan-2-amine: a potent, the disease remains to be established. For 15. Deacon CF, Wamberg S, Biev P, Hughes orally active dipeptidyl peptidase IV in- further studies, it is important to evaluate TE, Holst JJ: Preservation of active incre- hibitor for the treatment of type 2 diabe-

DIABETES CARE, VOLUME 30, NUMBER 6, JUNE 2007 1349 DPP-4 inhibitors

tes. J Med Chem 48:141–151, 2005 cacy and tolerability of the dipeptidyl 541, 2006 25. Ahre´n B: Vildagliptin: an inhibitor of peptidase-4 inhibitor sitagliptin as mono- 45. Rosenstock J, Brazg RG, Andryuk PJ, Lu dipeptidyl peptidase-4 with antidiabetic therapy over 12 weeks in patients with K, Stein P: Efficacy and safety of the properties. Exp Opin Invest Drugs 15:431– type 2 diabetes. Int J Clin Pract 61:171– dipeptidyl peptidase-4 inhibitor sitaglip- 442, 2006 180, 2007 tin added to ongoing pioglitazone therapy 26. Ahre´n B, Landin-Olsson M, Jansson PA, 35. Hanefeld M, Herman G, Mickel C, in patients with type 2 diabetes: a 24- Svensson M, Holmes D, Schweizer A: In- McGowan A, Wu M: Zhao P, Stein P. Ef- week, multicenter, randomized, double- hibition of dipeptidyl peptidase-4 re- fect of MK-0431, a dipeptidyl peptidase blind, placebo-controlled, parallel-group duces glycemia, sustains insulin levels IV (DPP-IV) inhibitor, on glycemic con- study. Clinical Ther 28:1556–1568, 2006 and reduces glucagon levels in type 2 di- trol after 12 weeks in patients with type 2 46. Matikainen N, Ma¨ntta¨ri S, Schweizer A, abetes. J Clin Endocrinol Metab 89:2078– diabetes. Diabetelogia 49 (Suppl. 1):A287, Ulvestad A, Mills D, Dunning BE, Foley 2084, 2004 2005 JE, Taskinen MR: Vildagliptin therapy re- 27. Herman GA, Bergman A, Stevens C, Kotey 36. Raz I, Hanefeld M, Xu L, Caria C, Wil- duces postprandial intestinal triglyceride- P, Yi B, Zhao P, Dietrich B, Golor G, liams-Herman D, Khatami H: Efficacy and rich lipoprotein particles in patients with Schrodter A, Keymeulen B, Lasseter KC, safety of the dipeptidyl peptidase-4 inhib- type 2 diabetes. Diabetologia 49:2049– Kipnes MS, Snyder K, Hilliard D, Tanen itor sitagliptin as monotherapy in patients 2057, 2006 M, Cilissen C, De Smet M, de Lepeleire I, with type 2 diabetes mellitus. Diabetologia 47. Hansotia T, Baggio LL, Delmeire D, Hinke van Dyck K, Wang AQ, Zeng W, Davies 49:2564–2571, 2006 SA, Yamada Y, Tsukiyama K, Seino Y, MJ, Tanaka W, Holst JJ, Deacon CF, 37. Aschner P, Kipnes MS, Lunceford JK, Holst JJ, Schuit F, Drucker DJ: Double in- Gottesdiener KM, Wagner JA: Effect of Sanchez M, Mickel C, Williams-Herman cretin receptor knockout (DIRKO) mice single oral doses of sitagliptin, a dipepti- DE: Effect of the dipeptidyl peptidase-4 reveal an essential role for the enteroinsu- dyl peptidase-4 inhibitor, on incretin and inhibitor sitagliptin as monotherapy on lar axis in transducing the glucoregula- plasma glucose levels after an oral glucose glycemic control in patients with type 2 tory actions of DPP-IV inhibitors. Diabetes tolerance test in patients with type 2 dia- diabetes. Diabetes Care 29:2632–2637, 53:1326–1335, 2004 betes. J Clin Endocrinol Metab 91:4612– 2006 48. Ahre´n B, Pacini G, Foley JE, Schweizer A: 4619, 2006 38. Nonaka K, Kakikawa T, Sato A, Okuyama Improved meal-related ␤-cell function 28. Bergman AJ, Coyte J, Yi B: Effect of renal K, Fujimoto G, Hayashi N, Suzuki H, and insulin sensitivity by the dipeptidyl insufficiency on the pharmacokinetics of Hirayama Y, Stein P: Twelve-week effi- peptidase-IV inhibitor vildagliptin in MK-0431 (sitagliptin), a selective dipep- cacy and tolerability of sitagliptin, a metformin-treated patients with type 2 tidyl-peptidase-IV (DPP-IV) inhibitor dipeptidyl peptidase-IV (DPP-4) inhibi- diabetes over 1 year. Diabetes Care 28: (Abstract). Clin Pharmacol Therapeut 38: tor, in Japanese patients with T2DM (Ab- 1936–1940, 2005 PII-46, 2006 stract). Diabetes 55 (Suppl. 1):A128, 2006 49. Balas B, Baig MR, Watson C, Dunning BE, 29. Ristic S, Byiers S, Foley J, Holmes D: Im- 39. Ahre´n B, Gomis R, Standl E, Mills D, Ligueros-Saylan M, Wang Y, He YL, Dar- proved glycaemic control with dipeptidyl Schweizer A: Twelve- and 52-week effi- peptidase-4 inhibition in patients with cacy of the dipeptidyl peptidase IV inhib- land C, Holst JJ, Deacon CF, Cusi K, Mari type 2 diabetes: vildagliptin (LAF237) itor LAF237 in metformin-treated A, Foley JE, DeFronzo RA: The dipeptidyl dose response. Diabet Obes Metab 7:692– patients with type 2 diabetes. Diabetes peptidase IV inhibitor vildagliptin sup- 698, 2005 Care 27:2874–2880, 2004 presses endogenous glucose production 30. Pratley RE, Jauffret-Kamel S, Galbreath E, 40. Bosi E, Camisaca RP, Collober C, Roch- and enhances islet function after single Holmes D: Twelve-week monotherapy otte E, Garber AJ: Effects of vildagliptin on dose administration in type 2 diabetic pa- with the DPP-4 inhbitor vildagliptin im- glucose control over 24 weeks in patients tients. J Clin Endocrinol Metab 92:1249– proves glycemic control in subjects with with type 2 diabetes inadequately con- 1255, 2007 type 2 diabetes. Horm Metab Res 387: trolled with metformin. Diabetes Care 30: 50. Dunning BE, Foley JE, Ahre´n B: 423–438, 2006 890–895, 2007 function in health and disease: influence 31. Pi-Sunyer FX, Schweizer A, Mills D, De- 41. Nathwani A: The use of vildagliptin for of glucagon-like peptide-1. Diabetologia jager S: Efficacy and tolerability of vilda- treatment of patients with type 2 diabetes 48:1700–1713, 2005 gliptin monotherapy in drug-naı¨ve (Abstract). Presented at the 66th Scientific 51. Azuma K, Radikova Z, Mancino JM, To- patients with type 2 diabetes. Diabetes Res Sessions of the American Diabetes Associ- ledo FGS, Thomas E, Lubowsky ND, Clin Pract 76:132–138, 2007 ation, Washington, DC, 9–13 June 2006 Kelkley CA, Kangani CO, Serra D, He Y, 32. Rosenstock J, Baron MA, Dejager S, Mills 42. Fonseca V, DeJager S, Albrecht D, Shirt L, Ligueras-Saylan M, Foley JE, Kelley DE: D, Schweizer A: Comparison of vildaglip- Schweizer A: Vildagliptin as add-on to in- DPP-4 inhibition improves insulin resis- tin and rosiglitazone monotherapy in pa- sulin in patients with type 2 diabetes tance (IR) in type 2 DM (Abstract). Pre- tients with type 2 diabetes: a 24-week, (T2D). Diabetes 55 (Suppl 1.):A111, 2006 sented at the 66th Scientific Sessions of double-blind, randomized trial. Diabetes 43. Charbonnel B, Karasik A, Liu J, Wu M, the American Diabetes Association, Care 30:217–223, 2007 Meininger G: Efficacy and safety of the Washington, DC, 9–13 June 2006 (Late- 33. Dejager S, Lebeaut A, Couturier A, dipeptidyl peptidase-4 inhibitor sitaglip- breaking abstract 5-LB) Schweizer A: Sustained reduction in A1C tin added to ongoing metformin therapy 52. Vella A, Bock G, Giesler PD, Burton DB, during one-year treatment with vildaglip- in patients with type 2 diabetes inade- Rizza RA, Camilleri M: The effect of the tin in patients with type 2 diabetes quately controlled with metformin alone. dipeptidyl peptidase IV inhibitor LAF237 (T2DM) (Abstract). Diabetes 55 (Suppl. Diabetes Care 29:2638–2643, 2006 on gastrointestinal function and glucose 1):A29, 2006 44. Jago C: American Diabetes Association metabolism in type 2 diabetes. Diabetolo- 34. Scott R, Wu L, Sanchez M, Stein P: Effi- 66th Scientific Sessions. IDrugs 9:538– gia 49 (Suppl. 1):26–27, 2006

1350 DIABETES CARE, VOLUME 30, NUMBER 6, JUNE 2007