Effect of Dipeptidyl Peptidase- 4 (DPP-4) Inhibitor on Biomarkers of Kidney Injury and Vascular Calcication in Diabetic Nephropathy: A Randomized Controlled Trial

Introduction: Dipeptidyl peptidase-4 (DPP-4) inhibitors improve glycemic control and have pleiotropic effects on kidney injury, albuminuria and vascular inammation, especially in animal models. We plan to evaluate the effects of a potent DPP4 inhibitor (gemigliptin) on these processes in diabetic nephropathy patients. Methods: This was a multicenter, prospective, randomized, placebo-controlled trial. A total of 201 participants were enrolled and randomly assigned to a group treated with 50 mg gemigliptin daily with standard care of diabetes mellitus for 6 months. The changes in coronary calcium score (CAC score), cardio-ankle vascular index (CAVI), estimated glomerular ltration rate (GFR), vascular calcication and tubular renal injury markers were evaluated at baseline and 6 months. Results: In total, 182 patients completed the study. Signicant reductions in hemoglobin A1C levels were observed in both groups. The changes in CAC score, CAVI, estimated GFR and proteinuria over the 6 months of the study did not signicantly differ between the groups. However, biomarkers of vascular calcication, including serum bone alkaline phosphatase, and kidney injury, including urine neutrophil gelatinase-associated lipocalin (NGAL)/Cr and urine liver fatty acid-binding protein (L-FABP)/Cr, were improved signicantly in the gemigliptin treatment group compared to the control group. No serious adverse event was observed during the study. Conclusion: Our study shows that gemigliptin signicantly improves renal tubular injury biomarkers and vascular calcication in patients with diabetic nephropathy; however, gemigliptin does not affect renal function or coronary calcication compared with the control. A larger and longer follow-up will be essential to determine these benecial effects. assay. a centrifuged particulate matter, and stored at -20°C before assay. Biochemistry indexes, including complete blood count, blood urea nitrogen (BUN), creatinine, fasting plasma glucose (FPG), hemoglobin A1c (HA1C), calcium, phosphorus, albumin and 24-hour urine protein, were measured at baseline and at each study visit. The coronary artery calcium (CAC) score; cardio-ankle vascular index (CAVI); vascular calcication biomarkers including serum osteopontin, bone alkaline phosphatase and reactive oxygen species (myeloperoxidase); and renal injury biomarkers including urine kidney injury molecule-1 (KIM-1), neutrophil gelatinase-associated lipocalin (NGAL) and liver fatty acid-binding protein (L-FABP) were determined at baseline and after 6 months. All vascular and renal injury biomarkers were tested by the quantitative sandwich enzyme immunoassay technique (R&D Systems, USA) according to the manufacturer’s instructions. Absorbance was measured at 450-570 nm using a microplate reader (SunriseTM Absorbance Reader, TECAN, Switzerland). Urine creatinine was assayed by the enzymatic method (Archetect C16000 analyzer, Tamecula, California). 8 Agatston gemigliptin treatment. To our knowledge, this is the rst study to show that DPP4 inhibitors alleviate kidney injury through the use of new biomarkers. Urine NGAL has been shown to be increased in many pathologic conditions, including diabetic kidney disease. 16 The level of urinary NGAL appears to increase beginning in the early phase of diabetic nephropathy, and the level is independently associated with albuminuria. 16,17 The novel nding of this study is that urine NGAL was signicantly decreased after gemigliptin treatment, which may indicate that gemigliptin helps ameliorate tubulointerstitial damage. Indeed, the evolution of the estimated GFR was not signicantly changed. A short follow-up time may inuence the therapeutic response. NGAL is a biomarker to monitor L-FABP regulates fatty acid uptake and intracellular transport, and the excretion rate of this protein is associated with tubulointerstitial structural damage. Therefore, L-FABP can be used to identify patients with a high susceptibility to renal stress. Increased urine concentrations of L-FABP have been observed in diabetic nephropathy. Indeed, urine L-FABP was elevated at the very early stage of diabetic nephropathy, even before any clinical signs of glomerular damage were detectable, and tubular damage, albuminuria and end-stage renal disease were independently predicted. 20 In this study, urine L-FABP markedly decreased after gemigliptin treatment. These results suggest that gemigliptin could reduce kidney injury and thus reduce the rate of renal disease progression. Long-term prospective follow-up may demonstrate whether eGFR is affected.


Background
The advantage of dipeptidyl peptidase-4 (DPP-4) inhibitors is due to their lower risk of inducing hypoglycemia in type 2 diabetic patients. 1 DPP-4 inhibitor action inhibits the proteolytic enzyme DPP-4, resulting in postponement of the degradation of glucagon-like peptide I (GLP-1), thus improving glycemic control, and GLP-1 also regulates calci cation of vascular smooth muscle cells through numerous pathways. 2 Gemigliptin is a potent DPP-4 inhibitor that has been approved for use in patients with type 2 diabetes, and it has pleiotropic effects in addition to its glucose-lowering effect, including inhibition of lipopolysaccharide (LPS)-induced proin ammatory effects in vascular endothelial cells by attenuating NF-kappa B and c-Jun NH (2)-terminal kinase (JNK) signaling via an AMP-activated protein kinase (AMPK)-dependent mechanism. 3 A recent study found that gemigliptin attenuated the calci cation of the abdominal aorta as well as RUNX2 and phosphate-induced Pit-1 mRNA expression, and reactive oxygen species formation. Therefore, gemigliptin could bene t vascular calci cation in patients with a high risk of disease, especially diabetic nephropathy.
DPP-4 inhibitors can be used safely in type 2 diabetic patients with renal impairment. Moreover, in an experimental model, DPP-4 inhibitors, especially gemigliptin, substantially decreased albuminuria and renal brosis in mice with unilateral ureteral obstruction 5 and attenuated podocyte injury in mice with diabetic nephropathy. 6 Recently, treatment with DPP-4 inhibitors attenuated kidney injury and improved acute and chronic kidney injury. 7 We therefore are interested in further investigating whether gemigliptin might have a similar effect in patients with diabetic nephropathy who have a high level of vascular calci cation and renal progression. We investigated the effect of gemigliptin on vascular calci cation and renal injury in type 2 diabetic patients with renal involvement.

Study population
We prospectively enrolled patients from three hospitals in Bangkok, Thailand: Vajira Hospital, Police Hospital and Phramongkutklao Hospital. The study implementation and protocol were approved by the institutional review board and adhered to the Declaration of Helsinki. Informed consent was obtained from all the participants before enrollment. The inclusion criteria were type 2 diabetic patients with stage 3-4 CKD (estimated glomerular ltration rate (GFR) 15-60 ml/min/1.73 mL), persistent micro-or macroalbuminuria and stable glycemic control for 12 weeks. The exclusion criteria included a history of allergy to DPP4 inhibitors, concurrent infectious disease, in ammatory diseases, post kidney transplantation, and treatment with calcimimetic agents, bisphosphonates, GLP-1 receptor agonists, DPP4 inhibitors and sodium-glucose cotransporter-2 inhibitors. The criteria to withdraw were patients who had other serious side effects after inclusion, severe hypoglycemia and needed hospitalization.

Study design
This was a multicenter prospective, open-label randomized controlled trial. Patients were randomized by the study coordinator into blocks of four, the allocation was concealed, and the patients were then divided into two groups at a ratio of 1:1. A computer-generated randomization procedure in blocks of four was used. The patients were then randomized into two groups: Group 1 received 50 mg gemigliptin per day for 6 months in addition to standard treatment, and Group 2 received standard treatment for type 2 diabetes and chronic kidney disease (CKD). Patients were scheduled for follow-up visits at months 2, 4 and 6, as shown in Figure 1. To give 80% power at P<0.05 to detect a difference in biomarker levels of vascular calci cation after gemigliptin treatment in CKD patients, a total of 100 patients per group were required, and 182 subjects total were recruited. 4 From September 2018 to October 2020, we collected baseline data from all participants for demographic characteristics, comorbid conditions and physical examination. Blood samples were collected and allowed to clot for 30 min at room temperature before centrifugation for 15 minutes. The serum was stored at -20°C before assay. Urine was collected in a sterile container, centrifuged to remove particulate matter, and stored at -20°C before assay. Biochemistry indexes, including complete blood count, blood urea nitrogen (BUN), creatinine, fasting plasma glucose (FPG), hemoglobin A1c (HA1C), calcium, phosphorus, albumin and 24-hour urine protein, were measured at baseline and at each study visit.
The coronary artery calcium (CAC) score; cardio-ankle vascular index (CAVI); vascular calci cation biomarkers including serum osteopontin, bone alkaline phosphatase and reactive oxygen species (myeloperoxidase); and renal injury biomarkers including urine kidney injury molecule-1 (KIM-1), neutrophil gelatinase-associated lipocalin (NGAL) and liver fatty acid-binding protein (L-FABP) were determined at baseline and after 6 months. All vascular and renal injury biomarkers were tested by the quantitative sandwich enzyme immunoassay technique (R&D Systems, USA) according to the manufacturer's instructions. Absorbance was measured at 450-570 nm using a microplate reader (SunriseTM Absorbance Reader, TECAN, Switzerland). Urine creatinine was assayed by the enzymatic method (Archetect C16000 analyzer, Tamecula, California).
Coronary artery calcium scoring (CAC score) All patients underwent computer tomography (CT) examination using an Ingenuity CT scanner (128 slice, Philips Medical Systems, Nederland B.V.) and were acquired with axial prospective gating while breath holding. The calcium score of each lesion was calculated by using the Agatston method. 8 The software package (Heart Beat Calcium Scoring, IntelliSpace Portal version 7.0; Philips Medical Systems, Nederland B.V.) automatically calculated and displayed the CT attenuation values. Agatston scores were reported for each of the 4 major coronary arteries, and the sum of these arteries was reported.

Cardio-ankle vascular index (CAVI)
We measured CAVI using a vascular screening system (Vasera VS-1500; Fukuda Denshi, Co. Ltd., Tokyo, Japan), with the patient lying supine and the head placed in the midline position. Cuffs were then applied bilaterally to the arms and ankles, and the patient was allowed to rest for 15 min. Measurements began by obtaining the blood pressure of the right brachial artery and ankle, followed by the left brachial artery and ankle. PWV was measured by dividing the vascular length by the time taken for the pulse wave to propagate from the aortic valve to the ankle. The ankle brachial pressure index (ABI) was also calculated.
CAVI was considered normal when the value was <8.

Statistical analysis
Continuous and categorical variables are described as the mean± standard deviation and numbers with parentheses. Differences between groups were compared using Student's t test and the chi-square test, respectively. The number and percentage of the variables are presented by treatment group, and differences between the two treatment groups were compared using Fischer's exact test. The comparison between variables before and after treatment with gemigliptin was performed using ANOVA. Multiple linear regression analysis was performed to evaluate the effects of HbA1c on renal injury markers. The level of signi cance for all statistical tests was set at 0.05. All analyses were performed using SPSS( Armonk ,NY:IBM Corp,USA) for Windows software version 22.

Patient characteristics
Of the 260 patients screened for eligibility, 201 (77%) patients were enrolled in this randomized clinical trial. Of the participants, 6 were withdrawn due to personal reasons, and 13 were excluded for other reasons. The remaining 182 participants completed the study, as shown in Figure 1.
There were 107 males and 75 females included in this study. The mean age was 62.77±9.59 years. Sixtytwo patients (80.52%) and 28 patients (36.36%) had hypertension and dyslipidemia, respectively. A total of 14.29% (n=11) had a previous history of ischemic heart diseases, while 1 patient (1.3%) had a history of peripheral arterial disease. At baseline, the 182 patients had a mean HbA1c of 8.25±1.83%, a mean estimated GFR of 45.58± 20.18 mL/min/1.73 m 2 and a mean body mass index of 28.16±5.37 kg/m 2 . The demographic and baseline characteristics of the groups were comparable, as shown in Table 1.  Table 4).

Gemigliptin and vascular calci cation
The calcium content in the coronary wall as measured by CAC scores increased over time in both groups but did not signi cantly differ before and after the treatment period (CAC score in the gemigliptin group increased from 655.72±905.88 to 690.93±932.89, P=0.18 and increased from 729.95±1123.98 to 744.67± 1148.90 in the control group). The change in CAC score was also nonsigni cant between the two groups ( Table 4).

Gemigliptin and vascular stiffness
We measured CAVI to assess arterial stiffness. After treatment, CAVI tended to be improved in the gemigliptin group (9.37±1.35 at baseline vs. 9.08±1.52 at 6 months, P=0.08), and CAVI was signi cantly improved the control group (9.26±1.44 at baseline vs. 8.69±2.06 at 6 months, P=0.005). However, the change in CAVI was not different in the two groups (P=0.265) ( Table 4 and Figure 3).

Gemigliptin and markers of vascular calci cation
To examine whether gemigliptin has protective effects against vascular calci cation, we examined the biochemical markers involved in vascular calci cation and oxidative stress. Serum osteopontin showed no signi cant differences from baseline levels in both groups, and the mean changes were not signi cantly different between the two groups ( Table 4). The major bone mineralization regulator (bone alkaline phosphatase) decreased in the gemigliptin group but increased in the control group. At 6 months, the level of bone alkaline phosphatase was signi cantly reduced in the gemigliptin treatment group compared to the control group (-5.84± 10.65 µg/L vs. 0.08±11.45 µg/L, P<0.001, respectively) ( Table 4 and Figure 4). However, serum myeloperoxidase levels, which indicate oxidative stress, did not change signi cantly from baseline, and the mean changes did not differ between the two groups.

Gemigliptin, renal fnction and proteinuria
To verify the short-term effect of gemigliptin on estimated GFR and proteinuria, the changes in estimated GFR from a point of treatment to 6 months were compared with the standard control group. The mean changes in estimated GFR and urine protein were not signi cantly different from baseline in the gemigliptin and control groups.

Gemigliptin and renal injury biomarkers
We examined the effects of gemigliptin on the levels of urinary renal injury biomarkers, such as NGAL, L-FABP and KIM-1. We adjusted urine biomarker concentrations by urine creatinine levels to eliminate the effects of patient hydration status. Urine NGAL tended to decrease but did not reach statistical signi cance in the gemigliptin group (387.9±1094.02 ng/mg creatinine at baseline vs. 316.37±679.96 ng/mg creatinine at the end of the study, P=0.412) ( Table 2), but urine NGAL signi cantly increased in the control group (333.67 ± 627.23 ng/mg creatinine at baseline vs. 590.95 ± 1252.05 ng/mg creatinine at the end of the study, P= 0.024) ( Table 3). The change in urine NGAL between the two groups was signi cantly different (-71.53 ± 837.30 ng/mg creatinine in the gemigliptin group vs 257.28 ± 1,047.29 ng/mg creatinine in the control group, P=0.020) ( Table 4 and Figure 5).
Urine LFABP decreased signi cantly in the gemigliptin group (91.32 ± 146.92 µg/mg creatinine at baseline vs. 37.16±68.47 µg/mg creatinine at the end of study, P<0.001), while urine LFABP was not signi cantly changed (48.86±67.14 µg/mg creatinine at baseline vs. 55.47 ± 86.31 µg/mg creatinine at the end of study, P=0.488). The degree of change in the control group was signi cantly different from that in the gemigliptin group (-54.17 ± 141.64 µg/mg creatinine in the gemigliptin group vs 6.6 ± 88.94 µg/mg creatinine in the control group, P=0.001) ( Table 4 and Figure 6). The effect of HbA1c on lowering renal biomarkers was not signi cant according to multiple linear regression analysis (Supplemental le).
Urine KIM-1 decreased in both groups, but the change in urine KIM-1 did not differ between the groups (-0.28±1.15 ng/mg creatinine in the gemigliptin group vs -0.50± 0.90 ng/mg creatinine in the control group, P=0.156). (Figure 7) The comparison of signi cant parameters are shown in supplemental le Adverse events Adverse events due to gemigliptin was rare. The most commonly reported adverse events from previous data were hypoglycemia, upper respiratory tract infection, urinary tract infection, nasopharyngitis, headache, diarrhea, arthralgia, hypertension, and cough. 9 However, we found only one case that reported palpitation after gemigliptin administration. Finally, the patient asked for withdrawal from the trial.

Discussion
Many investigators have studied the pleiotropic properties of DPP-4 inhibitors to highlight their potential bene ts in various diseases. At present, there are scarce data evaluating the effects of DPP-4 inhibitors on vascular calci cation in vivo. We chose gemigliptin due to its unique characteristics, including its action as a highly competitive and selective DPP-4 inhibitor. Then, we evaluated the effect of gemigliptin on vascular calci cation by using CAC scores and CAVI. After a follow-up period of 6 months, the CAC score and CAVI were not different between the two groups. This may be because the period of observation was too short. However, the markers of vascular calci cation and bone alkaline phosphatase (BALP) decreased signi cantly in the gemigliptin group compared with the control group. BALP is a sensitive and speci c marker for osteoblast activity and bone formation. Alkaline phosphatase stimulates mineralization mainly through the modulation of the balance between inorganic phosphate and inorganic pyrophosphate and has a role in cardiovascular remodeling. Yan et al. 10 showed that BALP was an independent risk factor for abdominal aortic calci cation and suggested a strong relationship between BALP and vascular calci cation in a dialysis population. Shantouf et al. 11 reported a signi cant association between serum alkaline phosphatase and coronary artery calci cation in maintenance hemodialysis. Taken together, the decrease in BALP and tendency to increase osteopontin levels by gemigliptin in our study might suggest the potential role of DPP4 inhibitors in protecting against vascular calci cation in the long term.
The role of DPP-4 inhibitors in renal disease is not fully understood. DPP-4 is highly expressed in proximal renal tubular cells and has proteolytic activity by involving the extracellular catabolism of proteins in the kidney, such as proline-containing peptide. DPP-4 inhibition is likely to alter the degradation/regulation of peptides in the lumen and thus in uence tubular structure and function in diabetes. 12 In addition to the glucose-lowering effects of DPP-4 inhibitors and the tissue-protective effects of DPP-4 inhibition have been demonstrated in particular ischemia-reperfusion injury, diabetic kidney disease and CKD. Kim et al. 13 reported that gemigliptin treatment led to a reduction in apoptosis, in ammation, and oxidative stress in a murine model of adriamycin-induced nephropathy. Choi et al. 14 showed that gemigliptin attenuated cisplatin-induced renal dysfunction in mice. The mechanisms were possibly due to inhibition of the apoptotic death of renal tubular cells and in ammatory responses. Of interest, we studied biomarkers of early kidney injury, such as KIM-1, NGAL and LFABP, which are not only more sensitive than serum creatinine for the identi cation of acute kidney injury but can also indicate speci c damage to the proximal tubule. 15 Urine NGAL and urine LFABP were signi cantly decreased after gemigliptin treatment. To our knowledge, this is the rst study to show that DPP4 inhibitors alleviate kidney injury through the use of new biomarkers. Urine NGAL has been shown to be increased in many pathologic conditions, including diabetic kidney disease. 16 The level of urinary NGAL appears to increase beginning in the early phase of diabetic nephropathy, and the level is independently associated with albuminuria. 16,17 The novel nding of this study is that urine NGAL was signi cantly decreased after gemigliptin treatment, which may indicate that gemigliptin helps ameliorate tubulointerstitial damage.
Indeed, the evolution of the estimated GFR was not signi cantly changed. A short follow-up time may in uence the therapeutic response. NGAL is considered a biomarker to monitor disease progression.
L-FABP regulates fatty acid uptake and intracellular transport, and the excretion rate of this protein is associated with tubulointerstitial structural damage. 18 Therefore, L-FABP can be used to identify patients with a high susceptibility to renal stress. Increased urine concentrations of L-FABP have been observed in diabetic nephropathy. 19 Indeed, urine L-FABP was elevated at the very early stage of diabetic nephropathy, even before any clinical signs of glomerular damage were detectable, and tubular damage, albuminuria and end-stage renal disease were independently predicted. 20 In this study, urine L-FABP markedly decreased after gemigliptin treatment. These results suggest that gemigliptin could reduce kidney injury and thus reduce the rate of renal disease progression. Long-term prospective follow-up may demonstrate whether eGFR is affected.
KIM-1 is a transmembrane protein that is markedly upregulated at the renal proximal tubule after injury. 21 Urinary levels of KIM-1 were signi cantly elevated in diabetic patients, indicating the existence of diabetic tubular damage at the early stage of diabetic nephropathy. 22 The expression of KIM-1 is mainly upregulated in proximal tubule cells both in rodents and in humans. 23 We found that urine KIM-1 was decreased signi cantly in both groups. The difference in urine KIM-1 was similar in both groups. However, KIM-1 may not be as sensitive of a marker as urine LFABP and NGAL to detect the favorable effects of gemigliptin. The mechanism by which gemigliptin signi cantly reduces renal injury may be related to the anti brotic, antiapoptotic and anti-in ammatory action of DPP4 inhibitors independent of their antiproteinuric and glucose-lowering effects. The markers that we chose to study are sensitive and established indicators for subtle insults to the kidney before measurable functional decline. These urinary markers were elevated before microalbuminuria was observed; thus, their values might be altered due to the changes in estimated GFR.
There were some limitations to our study. First, patients were followed for a relatively short duration (6 months). The antiatherosclerotic effects of gemigliptin, such as the effects on the CAC score and CAVI, could not be clearly demonstrated. The exposure to the study drug may not have been long enough to reverse the effects of years of pro-atherosclerotic processes in patients with a median duration of diabetes mellitus of more than 10 years; therefore, the study does not exclude the possibility of either bene t or risk with a longer duration of gemigliptin therapy. A longer study period needs to be continued. Second, it is not surprising that DPP-4 inhibitors signi cantly lower glucose levels. However, we used multiple logistic regression analysis to adjust for this factor.
In conclusion, our study demonstrated that gemigliptin improved glycemic control, vascular calci cation markers and kidney injury biomarkers. Urinary excretion of these markers is an early, sensitive and speci c marker for diabetic nephropathy that helps denote the bene cial effect of gemigliptin. In addition to the rare side effects of this class of DPP-4 inhibitors, their pleiotropic actions help retard tubular injury and diabetic nephropathy progression, which is of great clinical relevance. Additional large and long-term studies are needed to con rm their clinical bene t and utility.

Declarations
The study and protocol were approved by the Navamindradhiraj institutional review board and adhered to the Declaration of Helsinki. Informed consent was obtained from all the participants before enrollment.The authors approved for the publication in this journal.

Competeing interestes
The authors declare no competing interests.
Author's contribution TT Design the protocol,collect data,wrote paper,conceptualize BS set up database,analyzed data,wrote paper SS collect data TC collect data AS Interpretation of CAVI data TS Interpretation of CAVI data All authors read and approved the nal manuscript. Tables   Table 1 Baseline  Data are presented as the mean ± SD or percentage. Data are presented as the mean ± SD.
CAVI, cardio-ankle vascular index; LFABP, liver acid-binding protein; NGAL, neutrophil gelatinaseassociated lipocalin; Kim-1, kidney injury molecule-1. CAVI, cardio-ankle vascular index; LFABP, liver acid-binding protein; NGAL, neutrophil gelatinaseassociated lipocalin; Kim-1, kidney injury molecule-1. CAVI, cardio-ankle vascular index; LFABP, liver acid-binding protein; NGAL, neutrophil gelatinaseassociated lipocalin; Kim-1, kidney injury molecule-1.   Changes in Bone alkaline phosphatase in both groups at baseline and at the end of study Figure 5 Comparisons between urine NGAL and urine NGAL factor for urine creatinine in both groups at baseline and at the end of study  Comparisons between urine LFABP and urine LFABP factor for urine creatinine in both groups at baseline and at the end of study Figure 7 Comparisons between urine KIM-1 and urine KIM-1 factor for urine creatinine in both groups at baseline and at the end of study

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupplementalFileDPP4.docx