**Urinary Excretion of** *N***1-Methylnicotinamide, as a Biomarker of Niacin Status, and Mortality in Renal Transplant Recipients**

**Carolien P.J. Deen 1,2,3,\*, Anna van der Veen 2, Martijn van Faassen 2, Isidor Minovi´c 2, António W. Gomes-Neto 1,4, Johanna M. Geleijnse 5, Karin J. Borgonjen-van den Berg 5, Ido P. Kema <sup>2</sup> and Stephan J.L. Bakker 1,3,4**


Received: 11 October 2019; Accepted: 11 November 2019; Published: 12 November 2019

**Abstract:** Renal transplant recipients (RTR) commonly suffer from vitamin B6 deficiency and its functional consequences add to an association with poor long-term outcome. It is unknown whether niacin status is affected in RTR and, if so, whether this affects clinical outcomes, as vitamin B6 is a cofactor in nicotinamide biosynthesis. We compared 24-h urinary excretion of *N*1-methylnicotinamide (*N*1-MN) as a biomarker of niacin status in RTR with that in healthy controls, in relation to dietary intake of tryptophan and niacin as well as vitamin B6 status, and investigated whether niacin status is associated with the risk of premature all-cause mortality in RTR. In a prospective cohort of 660 stable RTR with a median follow-up of 5.4 (4.7–6.1) years and 275 healthy kidney donors, 24-h urinary excretion of *N*1-MN was measured with liquid chromatography-tandem mass spectrometry LC-MS/MS. Dietary intake was assessed by food frequency questionnaires. Prospective associations of *N*1-MN excretion with mortality were investigated by Cox regression analyses. Median *N*1-MN excretion was 22.0 (15.8–31.8) μmol/day in RTR, compared to 41.1 (31.6–57.2) μmol/day in healthy kidney donors (*p* < 0.001). This difference was independent of dietary intake of tryptophan (1059 ± 271 and 1089 ± 308 mg/day; *p* = 0.19), niacin (17.9 ± 5.2 and 19.2 ± 6.2 mg/day; *p* < 0.001), plasma vitamin B6 (29.0 (17.5–49.5), and 42.0 (29.8–60.3) nmol/L; *p* < 0.001), respectively. *N*1-MN excretion was inversely associated with the risk of all-cause mortality in RTR (HR 0.57; 95% CI 0.45–0.71; *p* < 0.001), independent of potential confounders. RTR excrete less *N*1-MN in 24-h urine than healthy controls, and our data suggest that this difference cannot be attributed to lower dietary intake of tryptophan and niacin, nor vitamin B6 status. Importantly, lower 24-h urinary excretion of *N*1-MN is independently associated with a higher risk of premature all-cause mortality in RTR.

**Keywords:** urinary excretion of *N*1-methylnicotinamide; kidney transplantation; mortality; niacin status; dietary intake; tryptophan; vitamin B3

#### **1. Introduction**

Kidney transplantation is the preferred treatment for end-stage renal disease in terms of survival, quality of life and costs [1,2]. Advances in transplantation medicine have lifted the 1-year patient survival higher than 90% [3]. While short-term patient outcomes are continuing to improve, the long-term posttransplant survival has remained largely unchanged over the past few decades [4]. Compared with the general population, renal transplant recipients (RTR) are at highly increased risk of premature mortality [5]. Improving perspectives relies on the management of modifiable factors that impact long-term outcome in RTR, of which nutrition is increasingly acknowledged [6,7].

Recently, we found that RTR commonly suffer from vitamin B6 deficiency and its functional consequences that add to an association with poor long-term outcomes [8]. As vitamin B6 is an essential cofactor of key enzymes involved in de novo biosynthesis of nicotinamide from tryptophan [9], niacin deficiency might be lurking in these patients as well. Nicotinamide, nicotinic acid, and nicotinamide riboside are collectively referred to as niacin or vitamin B3, and are precursors of the metabolically active NAD+. Besides dietary intake of pre-formed niacin, the so-called tryptophan-nicotinamide pathway is critical to maintaining niacin status [10]. Ongoing NAD<sup>+</sup> supply from its metabolic precursors, collectively referred to as "niacin equivalents", is required to provide reducing equivalents for energy metabolism and substrates of NAD<sup>+</sup> consuming enzymes [11]. NAD<sup>+</sup> is catabolized to *N*1-methylnicotinamide (*N*1-MN) through methylation of nicotinamide in the liver, and the 24-h urinary excretion of *N*1-MN is considered the most reliable biomarker of niacin status [12–14].

It is unknown whether niacin status is affected in RTR and, if so, whether this affects clinical outcomes. Hence, this study aims to compare 24-h urinary excretion of *N*1-MN in RTR with that in healthy kidney donors, in relation to dietary intake of tryptophan and niacin as well as vitamin B6 status, and to investigate whether niacin status is associated with the risk of premature all-cause mortality in RTR.

#### **2. Materials and Methods**

#### *2.1. Study Population*

This prospective study was conducted in a well-characterized, single-center cohort of 707 RTR (aged ≥18 years) with a functioning graft for at least 1 year who visited the outpatient clinic of the University Medical Center Groningen, Groningen, the Netherlands, between 2008 and 2011 [15–17]. As a control group, 367 healthy kidney donors were included who participated in a screening program before kidney donation. Signed informed consent was obtained from all participating subjects and the study protocol was approved by the institutional review board (METc 2008/186) adhering to the Declaration of Helsinki. Exclusion of subjects with missing biomaterial or niacin supplementation use left 660 RTR and 275 kidney donors eligible for statistical analyses (Figure S1).

#### *2.2. Data Collection*

All baseline measurements were obtained during a morning visit to the outpatient clinic. Participants were instructed to collect a 24-h urine sample on the day before their visit, and to fast overnight for 8 to 12 h. Urine samples were collected under oil, and chlorhexidine was added as an antiseptic agent. Fasting blood samples were drawn after completion of the urine collection. Blood was collected in a series of evacuated tubes with different additives (Vacutainer®; BD, Franklin Lakes, NJ, USA) for preparation of plasma and serum. Body composition and hemodynamic parameters were measured according to a previously described, strict protocol [15]. Serum parameters, including lipid, inflammation, and glucose homeostasis variables were measured with spectrophotometric-based routine clinical laboratory methods (Roche Diagnostics, Rotkreuz, Switzerland). Diabetes was diagnosed if fasting plasma glucose was ≥7.0 mmol/L or antidiabetic medication was used [15]. Plasma vitamin B6 was determined as its principal, metabolically active form pyridoxal-5 -phosphate using a

HPLC method (Waters Alliance, Milford, MA, USA) with fluorescence detection (JASCO, Inc., Easton, MD, USA) [8].

Renal function was assessed by estimation of the glomerular filtration rate (eGFR) and detection of proteinuria. The eGFR was calculated using the combined creatinine and cystatin C-based Chronic Kidney Disease Epidemiology Collaboration equation [18], which has been shown to be the most accurate equation in RTR [19]. Proteinuria was diagnosed if total urinary protein excretion was ≥0.5 g/day as measured by a biuret reaction-based assay (MEGA AU510; Merck Diagnostica, Darmstadt, Germany).

Dietary intake including tryptophan and niacin intakes was assessed with a validated semi-quantitative food frequency questionnaire (FFQ) [20–22]. The self-administered questionnaire was filled out at home and inquired about 177 food items during the last month, taking seasonal variations into account. During the visit to the outpatient clinic, the FFQ was checked for completeness by a trained researcher and inconsistent answers were verified with the participant. The FFQ was validated for RTR as previously reported [16]. Dietary data were converted into daily nutrient intake using the Dutch Food Composition Table of 2006 [23]. Alcohol consumption and smoking behavior were assessed with a separate questionnaire [6]. Additional data on medical history and use of medication and vitamin supplements were obtained from medical records [6].

#### *2.3. Assessment of N1-MN Excretion*

Measurement of *N*1-MN concentration was performed with a validated liquid chromatography (Luna HILIC column; Phenomenex, Torrance, CA, USA) isotope dilution-tandem mass spectrometry (LC-MS/MS) (Quattro Premier; Waters, Milford, MA, USA) method, as described previously [24]. The 24-h urinary excretion of *N*1-MN (μmol/day) was obtained after multiplying *N*1-MN concentration (μmol/L) by total urine volume calculated from weight (L/day). The reference range of *N*1-MN excretion in healthy individuals was previously established at 17.3–115 μmol/day [24].

#### *2.4. Clinical Endpoints*

The primary outcome of this study was all-cause mortality which was recorded until 30 September 2015 with no loss due to follow-up. RTR status was kept up-to-date through the continuous surveillance system of the outpatient program.

#### *2.5. Statistical Analysis*

Data are presented as the mean ± SD, median (IQR) and absolute number (percentage) for normally distributed, skewed, and nominal data, respectively. Assumptions for normality were checked by visual judgments of the corresponding frequency distribution and Q-Q plot.

Baseline characteristics of RTR and healthy kidney donors were compared by means of *t*, Mann-Whitney, and Chi-Square tests. Niacin status in RTR and healthy kidney donors was compared by linear regression analyses of 2-base log-transformed *N*1-MN excretion, with subsequent cumulative adjustment for age and sex (model 1), eGFR (model 2) and intake of energy, tryptophan, and niacin and plasma vitamin B6 (model 3).

RTR characteristics were divided into tertiles of *N*1-MN excretion stratified by sex (T1, T2, and T3) and compared by means of ANOVA, Kruskal-Wallis, and Chi-Square tests.

For prospective analyses, a Cox proportional hazards regression model for all-cause mortality outcome was fitted to *N*1-MN excretion as a sex-stratified tertile-based categorical variable, as well as a continuous variable adjusted for sex (model 1). Confounding was controlled for by including potential confounders as covariates in the regression model. Crude associations were adjusted cumulatively for age (model 2), smoking and body surface area (model 3) and, to prevent overfitting, additionally for intake of alcohol and energy and plasma vitamin B6 (model 4), kidney function (model 5), medication use (model 6), and high-sensitivity C-reactive protein (hs-CRP) (model 7). Variables that could lie in the causal pathway of *N*1-MN excretion and all-cause mortality were not adjusted for because this

might obscure otherwise existing associations unintentionally. Assumptions of proportionality of the hazard functions and the linearity of log-hazards were checked by visual judgements of Kaplan Meier plots of the survival and log-survival function entering the sex-stratified *N*1-MN excretion tertile group variable.

In secondary analyses, effect modification was assessed by including the cross product term of each potential confounder and 2-base log-transformed *N*1-MN excretion in the Cox regression model adjusted for age and sex (model 2). Subsequent stratified analyses were performed for subgroups of significant effect modifiers on the association of *N*1-MN excretion with all-cause mortality.

For all statistical analyses, a two-sided *p*-value of less than 0.05 was considered to indicate statistical significance and SPSS Statistics version 23.0 (IBM, Armonk, NY, USA) was used as software.

#### **3. Results**

#### *3.1. Baseline Characteristics and Comparison of N1-MN Excretion*

This study included 660 stable RTR (57% male; mean age 53.0 ± 12.7 years), at a median time of 5.6 (2.0–12.0) years after transplantation, and 275 healthy kidney donors (41% male; mean age 53.3 ± 10.7 years) (Table 1). Intake of tryptophan was similar in both groups (1059 ± 271 and 1089 ± 308 mg/day, respectively; *p* = 0.19), while intake of niacin was lower in RTR than in kidney donors (17.9 ± 5.2 and 19.2 ± 6.2 mg/day, respectively; *p* = 0.01). Taken together, intake of niacin equivalents was lower in RTR than in kidney donors (35.6 ± 9.2 mg/day and 37.4 ± 10.8, respectively; *p* = 0.03) (Figure 1). All RTR and kidney donors complied with the recommended daily intake that is set at a minimum of 6.6 niacin equivalents per 1000 kcal (≥ 9.6 and ≥ 11.7 mg/1000 kcal, respectively) [12]. As previously reported, RTR had significantly lower plasma vitamin B6 compared to kidney donors (29.0 (17.5–49.5) and 42.0 (29.8–60.3) nmol/L, respectively; *p* < 0.001). Median *N*1-MN excretion was 22.0 (15.8–31.8) μmol/day in RTR, compared to 41.1 (31.6–57.2) μmol/day in kidney donors (*p* < 0.001) (Figure 1). Furthermore, urinary excretion of *N*1-MN was below the reference limit of 17.3 μmol/day in 202 (31%) RTR, against 4 (2%) kidney donors. The difference in *N*1-MN excretion between RTR and kidney donors was independent of age, sex, eGFR, intake of energy, tryptophan, and niacin and plasma vitamin B6 (Table 2). Cyclosporine, azathioprine, and anticonvulsants were used by, respectively, 253 (38%), 112 (17%) of 19 (3%) of RTR, and none of the controls received drugs that are known to potentially affect niacin status.


**Table 1.** Baseline characteristics of stable RTR compared to that in healthy kidney donors 1.


**Table 1.** *Cont.*

<sup>1</sup> Data are presented as mean ± SD, median (IQR) and absolute number (percentage) for normally distributed, skewed and nominal data, respectively. <sup>2</sup> *p*-value for difference was tested by *t* and Mann-Whitney tests for normally and skewed distributed continuous variables, respectively, and Chi-Square tests for nominal variables. <sup>3</sup> Intake of niacin equivalents was calculated by adding up niacin and one-sixtieth of tryptophan intake. Subjects who were using niacin supplementation were excluded. eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c; *N*1-MN, *N*1-methylnicotinamide; RTR, renal transplant recipients.

**Figure 1.** Box plots of dietary intake of (**a**) tryptophan, (**b**) niacin and (**c**) niacin equivalents and (**d**) log2 24-h urinary excretion of *N*1-MN in RTR compared to that in healthy kidney donors. Boxes, bars and whiskers represent IQRs, medians and values <1.5 × IQR, respectively, whereas outliers (1.5–3 × IQR) are indicated by circles and extreme outliers (>3 × IQR) by asterisks. Log2 of the lower and upper bound of the reference range of *N*1-MN excretion in healthy individuals (17.3–115.0) μmol/day [24] are indicated with dotted lines (**d**). *p*-value for difference between RTR and donors was tested by t and Mann-Whitney tests for normally and skewed distributed continuous variables, respectively. Intake of niacin equivalents was calculated by adding up niacin and one-sixtieth of tryptophan intake. *N*1-MN, *N*1-methylnicotinamide; RTR, renal transplant recipients.


**Table 2.** Association of RTR and healthy kidney donors grouping with *N*1-MN excretion 1.

<sup>1</sup> Linear regression analyses were performed to investigate the association of RTR and healthy kidney donors as grouping variable with *<sup>N</sup>*1-MN excretion, with adjustment for potential confounders. <sup>2</sup> Model 1: crude model. <sup>3</sup> Model 2: adjusted for age and sex. <sup>4</sup> Model 3: adjusted as for model 2 and for eGFR. <sup>5</sup> Model 4: adjusted as for model 3 and for intake of energy, tryptophan and niacin and plasma vitamin B6. eGFR, estimated glomerular filtration rate; *N*1-MN, *N*1-methylnicotinamide; RTR, renal transplant recipients; std.β, standardized beta coefficient.

RTR characteristics across tertiles of sex-stratified *N*1-MN excretion (M: <19.2, 19.2–28.8, >28.8 μmol/day; F: <16.1, 16.1–25.6, >25.6 μmol/day in T1, T2, and T3, respectively) are shown in Table 3. Age and the presence of acetylsalicylic acid, proton pump inhibitors, diuretics and post mortem donors were lower with increasing tertiles of *N*1-MN excretion, while intake of alcohol, energy, tryptophan and niacin, plasma vitamin B6, kidney function and the presence of proliferation inhibitors and primary glomerular disease were higher with increasing tertiles of *N*1-MN excretion.




**Table 3.** *Cont.*

<sup>1</sup> Data are presented as mean ± SD, median (IQR) and absolute number (percentage) for normally distributed, skewed and nominal data, respectively. <sup>2</sup> *p*-value for difference was tested by ANOVA and Kruskal-Wallis tests for normally and skewed distributed continuous variables, respectively, and Chi-Square tests for nominal variables. eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c; hs-CRP, high-sensitivity C-reactive protein; *N*1-MN, *N*1-methylnicotinamide; RTR, renal transplant recipients.

Other or unknown cause, *n* (%) 39 (18) 36 (16) 35 (16) 0.85

#### *3.2. N1-MN Excretion and Mortality*

During a median follow-up time of 5.4 (4.7–6.1) years, 143 (22%) RTR died. The risk of all-cause mortality increased with lower tertiles of *N*1-MN excretion, as depicted by Kaplan-Meier curves (Figure 2). Cox regression analyses revealed an inverse association of *N*1-MN excretion with all-cause mortality (Model 2: HR 0.57; 95% CI 0.45–0.71; *p* < 0.001), independent of potential confounders (Table 4). The same held for analyses across tertiles of sex-stratified *N*1-MN excretion (Table 4). RTR in the lowest and middle tertiles were at higher risk of all-cause mortality compared to those in the highest tertile as reference (Model 2: HR 2.68; 95% CI 1.67–4.33; *p* < 0.001 and HR 2.04; 95% CI 1.25–3.34; *p* = 0.004, respectively), independent of potential confounders (Table 4).

**Figure 2.** Survival curves for all-cause mortality in RTR according to tertiles of sex-stratified *N*1-MN excretion. *N*1-MN excretion was <19.2, 19.2–28.8, and >28.8 μmol/day for males, and <16.1, 16.1–25.6 and >25.6 μmol/day for females in T1, T2, and T3, respectively. *N*1-MN, *N*1-methylnicotinamide; RTR, renal transplant recipients.

**Table 4.** Association of *N*1-MN excretion with risk of all-cause mortality in RTR 1.


<sup>1</sup> Cox regression analyses were performed to investigate the association of *N*1-MN excretion with risk of all-cause mortality in RTR, with adjustment for potential confounders. <sup>2</sup> *N*1-MN excretion was <19.2, 19. 2–28.8, and <sup>&</sup>gt;28.8 <sup>μ</sup>mol/day for males, and <sup>&</sup>lt;16.1, 16.1–25.6, and <sup>&</sup>gt;25.6 <sup>μ</sup>mol/day for females in T1, T2, and T3, respectively. <sup>3</sup> Model 1: not adjusted in tertiles of sex-stratified *<sup>N</sup>*1-MN excretion, adjusted for sex in continuous analyses. <sup>4</sup> Model 2: adjusted as for model 1 and for age. <sup>5</sup> Model 3: adjusted as for model 2 and for smoking and body surface area. <sup>6</sup> Model 4: adjusted as for model 3 and for intake of alcohol and energy and plasma vitamin B6. <sup>7</sup> Model 5: adjusted as for model 3 and for eGFR, proteinuria, donor status and primary glomerular disease. <sup>8</sup> Model 6: adjusted as for model 3 and for use of proliferation inhibitors, acetylsalicylic acid, proton pump inhibitors and diuretics. <sup>9</sup> Model 7: adjusted as for model 3 and for hs-CRP. eGFR, estimated glomerular filtration rate; hs-CRP, high-sensitivity C-reactive protein; *N*1-MN, *N*1-methylnicotinamide; RTR, renal transplant recipients.

Secondary analyses exposed significant effect modification of hs-CRP on the association of *N*1-MN excretion with all-cause mortality (*p* = 0.05), independent of age and sex. The inverse association of *N*1-MN excretion with all-cause mortality was stronger for individuals in the subgroup with serum hs-CRP <2.4 mg/L (HR 0.47; 95% CI 0.35–0.64; *p* < 0.001), than in the subgroup with serum hs-CRP ≥2.4 mg/L (HR 0.70; 95% CI 0.50–0.96; *p* = 0.03) according to subsequent stratified analysis.

#### **4. Discussion**

In this large prospective cohort study, we showed that RTR excrete less *N*1-MN in 24-h urine than healthy controls and our data suggest that this difference cannot be attributed to lower dietary intake of tryptophan and niacin, nor vitamin B6 status. Furthermore, lower 24-h urinary excretion of *N*1-MN as a biomarker of niacin status was independently associated with a higher risk of premature all-cause mortality in RTR.

To the best of our knowledge, niacin status has not been studied within the context of kidney transplantation and its concomitant long-term implications yet. In fact, prospective data on the urinary excretion of *N*1-MN have been limited to one previous study in patients recovering from leukemia treatment [25]. Studies on niacin nutrition in relation to prospective outcomes are likewise scarce, as the prevailing intake of niacin equivalents is suggested to be not sufficiently low to compromise survival. Presumed health benefits of niacin are pharmacological rather than physiological [26–29], although higher survival with higher niacin intake in elderly has been reported previously [30] in congruence with our findings.

Niacin is considered the least critical vitamin to meet the recommendations through dietary intake in western societies [31], as niacin equivalents are found in a wide range of foods [12]. In line with this, dietary intake of niacin equivalents was sufficient according to WHO guidelines in all RTR and healthy kidney donors, while we found that urinary excretion of *N*1-MN was commonly below the established reference bound in RTR. The observed disparity of *N*1-MN excretion between RTR and healthy kidney donors could moreover not be explained by lower dietary intake of niacin equivalents in RTR in the present study.

The fact that we found a positive association of plasma vitamin B6 concentration with *N*1-MN excretion strengthens our hypothesis that inadequacies of this cofactor might affect niacin status in RTR. Adjustment for plasma vitamin B6, however, neither did alter the discrepancy of *N*1-MN excretion between RTR and healthy kidney donors. Therefore, one should consider other factors that could interfere with *N*1-MN excretion as a biomarker of niacin status, and add to poor long-term outcome in RTR.

Whereas secondary dietary inadequacies may interrupt niacin metabolism, this also holds for certain medications including specific antituberculosis, anticonvulsant and antiproliferative drugs, as well as cyclosporine and azathioprine [32–34], which are common immunosuppressant drugs in RTR, although in our population those did not appear to affect *N*1-MN excretion.

We can furthermore speculate on the presence of enhanced consumption of tryptophan for protein biosynthesis at the cost of niacin status in RTR. Interestingly, tryptophan is argued to be quantitatively the most important NAD<sup>+</sup> precursor, as it is more effective in elevating liver NAD<sup>+</sup> and urinary excretion of *N*1-MN than the salvageable precursors [35–38]. The tryptophan-nicotinamide pathway is, however, mainly regulated by tryptophan intake rather than niacin status, since the generally accepted conversion ratio of 60:1 falls when dietary tryptophan is limiting [39]. Indeed, tryptophan is used primarily for protein biosynthesis and only after nitrogen balance has been achieved for the nicotinamide pathway [40]. This allows us to speculate on protein catabolism and negative protein balance as part of protein-energy wasting in RTR, engendered by metabolic derangement, systemic inflammation, acidemia, and the use of immunosuppressive drugs, to induce tryptophan consumption for protein synthesis in this population [41,42]. However, as our study was not designed to assess protein-energy wasting, we cannot conclusively address such an effect on *N*1-MN excretion in RTR.

On the contrary, the tryptophan-nicotinamide pathway is implicated in disease states in which systemic inflammation is present, by the enhanced action of indoleamine 2,3-dioxygenase in response to inflammatory cytokines and mediators. This upregulation of tryptophan degradation towards nicotinamide is known to yield relativity large amounts of quinolinic acid to fuel NAD+-consuming poly (ADP-ribose) polymerase (PARP) reaction in response to immune-related (oxidative) damage [35]. Although we observed lower serum hs-CRP levels as a low-grade inflammation biomarker with higher tertiles of *N*1-MN excretion, this difference did not reach significance.

Finally, the renal clearance of *N*1-MN itself can also be affected by several factors and not in the least by impaired kidney function. In fact, *N*1-MN is eliminated almost exclusively by the kidneys, being partly excreted partly by glomerular filtration and partly by tubular secretion with negligible and saturable tubular reabsorption [43]. Whereas renal clearance of *N*1-MN has been investigated as a model of renal secretory function [43] and to predict renal clearance of cationic drugs in renal insufficiency [44], plasma concentrations are suggested to be less sensitive to kidney function because of the contribution of aldehyde oxidase to *N*1-MN clearance, yielding *N*1-methyl-2-pyridone-5-carboxamide (2Py) [45]. Although our findings appeared independent of kidney function, future studies are warranted to rule out enhanced oxidative metabolism, causing a shift towards urinary excretion of 2Py in this population.

Regarding potential mechanisms for the association of *N*1-MN excretion with mortality, NAD<sup>+</sup> homeostasis has been linked to increased resistance against a range of pathophysiological processes that are predominant and impact poor long-term outcome in RTR, including cardiovascular, inflammatory, malignant and metabolic disorders [46]. The availability of NAD<sup>+</sup> is determined by its production from niacin equivalents, as well as its degradation in NAD<sup>+</sup> consuming activities [47]. NAD<sup>+</sup> levels remain constant when used as a coenzyme, being recycled back and forth between its oxidized and reduced forms [11], but are depleted by three distinct classes of enzymes that consume NAD<sup>+</sup> as a substrate: PARP, cyclic ADP ribose synthases (CD38 and CD157), and sirtuins [48]. Excessive activation of PARP and CD38 is induced by stresses such as inflammation, oxidative stress and DNA damage that are predominant in in RTR [48,49]. As a result, NAD<sup>+</sup> availability might become limiting for beneficial sirtuin activities; in particular with lower niacin status. These beneficial effects of sirtuins have been described more specifically for renal diseases, including renoprotective effects by inhibition of renal cell apoptosis, inflammation, and fibrosis and regulation of mitochondrial function and glucose, lipid, and energy metabolism [50–53].

Whereas we did not find an association of *N*1-MN excretion with hs-CRP, this low-grade inflammation biomarker appeared to affect the magnitude of the inverse association of *N*1-MN excretion with all-cause mortality. Although we can only speculate on the underlying mechanism, earlier mentioned inflammation-related overconsumption of NAD<sup>+</sup> limiting its downstream beneficial activities might at least partly explain the lower protective effect of niacin status on mortality in the subgroup with higher serum hs-CRP levels.

The current study should be interpreted within its strengths and limitations. First, its observational nature prohibits causal inferences, but it also did not allow us to draw conclusions on underlying mechanisms of lower *N*1-MN excretion in RTR and its contribution to worse survival. Second, the generalizability of our findings might be compromised by overrepresentation of Caucasian individuals from a single center, despite being controlled for by the inclusion of a large, representative control group. Third, the reliability of FFQ data is subject to sources of measurement error, including recall and social desirability biases and limitations in food composition databases [54]. Higher similarity in dietary sources could be achieved by including spouses as a control group. Finally, the present study is confined to the 24-h urinary excretion of *N*1-MN as the recommended biomarker of niacin nutritional status by authorities, including the WHO and the European Food Safety Authority [12–14]. Future studies are, however, encouraged to elaborate on plasma concentrations of niacin and its metabolites, or NAD<sup>+</sup> and the ratio of NAD<sup>+</sup> to NADP<sup>+</sup> in erythrocytes as additional indices of niacin status. Although observational evidence is inherent to limitations, prospective cohort studies provide a strong design to address nutritional status and health outcome associations over a long period of

time. Strengths of our study include its large sample size, with a sufficient number of incident cases and no loss to follow-up, and therefore minimizing the risk of bias in the assessment of outcome. The extensive characterization of participants, moreover, allowed us to control for confounding and effect modification in estimates of the effect.

#### **5. Conclusions**

In conclusion, 24-h urinary excretion of *N*1-MN as a biomarker of niacin status is lower in RTR than in healthy controls, and other factors than dietary intake of niacin equivalents and vitamin B6 status appear to reinforce this discrepancy. Importantly, 24-h urinary excretion of *N*1-MN is inversely associated with a higher risk of premature all-cause mortality in RTR and niacin status is therefore revealed as a potential target for nutritional strategies to improve long-term outcome after kidney transplantation. However, further research is warranted to unravel underlying mechanisms that potentially interfere with *N*1-MN excretion in RTR, and to strengthen causal inferences for health outcomes to support dietary recommendation.

**Supplementary Materials:** The following is available online at http://www.mdpi.com/2077-0383/8/11/1948/s1, Figure S1: Flow of participants through study protocol.

**Author Contributions:** The authors' responsibilities were as follows—S.J.L.B. and I.P.K.: designed research; M.v.F., A.W.G.-N., J.M.G. and K.J.B.-v.d.B.: provided essential materials; C.P.J.D. and A.v.d.V.: analyzed data; C.P.J.D. and S.J.L.B.: wrote paper and had primary responsibility for final content; A.v.d.V., M.v.F., I.M., A.W.G.-N. and J.M.G.: critically revised the manuscript for important intellectual content; and all authors: read and approved the final manuscript.

**Funding:** This research was funded by Top Institute Food and Nutrition, grant numbers A-1003 and 16NH01.

**Acknowledgments:** Supported by FrieslandCampina and Danone Nutricia Research. The cohort on which the study was based is registered at clinicaltrials.gov as "TransplantLines Food and Nutrition Biobank and Cohort Study (TxL-FN)" with number NCT02811835.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Validation of Identified Susceptible Gene Variants for New-Onset Diabetes in Renal Transplant Recipients**

**Hyeon Seok Hwang 1, Kyung-Won Hong 2, Jin Sug Kim 1, Yang Gyun Kim 1, Ju Young Moon 1, Kyung Hwan Jeong 1, Sang Ho Lee 1,\* and The Korean Organ Transplantation Registry Study Group**


Received: 9 September 2019; Accepted: 12 October 2019; Published: 16 October 2019

**Abstract:** Genome-wide association studies (GWAS) and candidate gene approaches have identified single nucleotide polymorphisms (SNPs) associated with new-onset diabetes after renal transplantation (NODAT).We evaluated associations between NODAT and SNPs identified in previous studies. We genotyped 1102 renal transplant recipients from the Korean Organ Transplantation Registry (KOTRY) database; 13 SNPs were assessed for associations with NODAT (occurring in 254 patients; 23.0%), within one year after transplantation. The frequency of the T allele at *KCNQ1* rs2237892 was significantly lower in patients with NODAT compared to control patients (0.30 vs. 0.39; *p* = 8.5 <sup>×</sup> 10−5). The T allele at rs2237892 was significantly associated with decreased risk of NODAT after adjusting for multiple variables, compared to the C allele (OR 0.63, 95% CI 0.51–0.79; *<sup>p</sup>* <sup>=</sup> 5.5 <sup>×</sup> <sup>10</sup><sup>−</sup>5). Dominant inheritance modeling showed that CT/TT genotypes were associated with a lower risk for development of NODAT (OR 0.56, 95% CI 0.42–0.76; *p* = 2.0 <sup>×</sup> 10−4) compared to the CC genotype. No other SNPs were associated with NODAT. Our study validated the protective effect of T allele at *KCNQ1* rs2237892 on the development of NODAT in a large cohort of renal transplant recipients. Our findings on susceptibility variants might be a useful tool to predict NODAT development after renal transplantation.

**Keywords:** new onset diabetes after renal transplantation; single nucleotide polymorphisms; renal transplantation

### **1. Introduction**

Development of new-onset diabetes after renal transplantation (NODAT) is a common complication in patients that have undergone transplantation. The cumulative incidence of NODAT is approximately 15%–30% at 1-year post-transplantation, and the annual incidence of NODAT is approximately 4%–6% [1–3]. This metabolic disorder induces a worse cardiovascular risk profile and results in a three-fold risk of cardiovascular morbidity [4,5]. In addition, NODAT is associated with a 1.5- to 3-fold risk of allograft loss and results in a 10%–20% reduction in long-term patient survival [1,6,7]. The accumulated health-care cost is also considerable, with an estimated cost of US \$21,500 per new patient with diabetes in the second year after transplantation [8]. Therefore, NODAT is a critical burden of recipient care and a major clinical challenge for the longevity and survival of renal allograft patients.

The risk of developing NODAT is associated with several clinical factors, including the recipient age, BMI, use of tacrolimus and corticosteroid, acute rejection, hepatitis C virus, cytomegalovirus infection, autosomal dominant polycystic kidney disease, and hypomagnesemia [1,9–15]. However, evidence suggests an increased incidence of NODAT despite the identification of clinical risk factors and the effort to mitigate the risk [16]. As current strategies have limited effectiveness in preventing NODAT, genetic risk stratification emerges as a key approach to address this problem.

Several studies have shown genetic predisposition as a risk factor for the development of NODAT. Genetic polymorphism studies on NODAT led to the identification of several candidate genes, derived from genome-wide association studies (GWAS) for type 2 diabetes [17,18]. Commonly evaluated genetic determinants included genes involving carbohydrate metabolism, insulin secretion, and insulin resistance [19]. In addition, genes that encode inflammatory cytokines correlated with type 2 diabetes and were also associated significantly with NODAT [20]. More recently, GWAS showed that genes involved in β-cell apoptosis are associated with the development of NODAT [21,22]. However, candidate gene approaches included only a few individuals with NODAT, leading to inconsistent results, and the significant genes identified in GWAS are not replicated in independent cohorts. Therefore, these limitations severely interrupt the development of prevention strategies against NODAT.

This study aimed to verify the association of previously identified genetic polymorphisms with NODAT in a large nationwide prospective cohort. We selected 17 single nucleotide polymorphisms (SNPs) on susceptibility loci and evaluated the effects of these independent SNPs on the risk of developing NODAT.

#### **2. Materials and Methods**

#### *2.1. Study Population*

The study population was selected from the Korean Organ Transplantation Registry (KOTRY), which is a prospective, multicenter, nationwide cohort study that includes transplantation information in Korea. Thirty-two representative national hospitals and transplantation centers participated in KOTRY. Recipients were enrolled consecutively upon undergoing a transplantation procedure and followed up accordingly from July 2014 to December 2018. The registry accumulated data on individual patients including demographics, comorbidities, laboratory data, induction and maintenance of the immunosuppressive regimen, and several other types of events. Our study was reviewed and approved by the Institutional Review Board of each transplantation center. All patients provided written informed consent before enrollment in the study.

Blood samples from 1826 patients were stored for genotyping and screened using the KOTRY database. The following patients were excluded: Renal transplant recipients with established diabetes (*n* = 503), patients followed up for less than one year (*n* = 107), non-functioning graft at one-year follow-up (*n* = 32), incomplete record of medical or laboratory findings (*n* = 65), missing information on human leukocyte antigen (HLA) typing (*n* = 2), and others (*n* = 15). In total, finally, 1102 patients were enrolled for this study.

#### *2.2. Selection of SNPs and Genotyping*

We conducted an extensive literature review for published variants that were significantly associated with NODAT in renal transplant recipients. We evaluated SNPs, which showed top-ranked associations with NODAT in individual studies. We selected seventeen SNPs that were significantly associated with NODAT from GWAS or well-established association studies of NODAT [18,20–22].

Blood samples (3 mL each) were collected in tubes containing RBC lysis solution. The blood sample from each study participant was centrifuged to obtain white blood cells. Genomic DNA was extracted from white blood cells using a DEXTM II genomic DNA extraction kit (Intron, Sungnam, Korea). DNA samples were stored at −80 ◦C before analysis. Quality of stored DNA samples was evaluated using agarose gel electrophoresis to confirm sample integrity. SNPs were genotyped from these DNA samples using TaqMan-based QuantStudio OpenArray® (Life Technologies, Carlsbad, CA, USA). DNA from patients and controls was randomly transferred into 96-well plates and genotyped using a blinded method. The call rates for genotyping of the SNPs were >98%.

#### *2.3. Data Collection and Definition*

We collected the following baseline patient characteristics at the time of transplantation: Age, gender, body mass index (BMI), relevant comorbid conditions, information on human leukocyte antigen (HLA), blood typing, desensitization, and induction and maintenance of the immunosuppressive regimen. Laboratory data were collected at baseline and regularly followed up. Clinical events were identified, including diabetes, the occurrence of biopsy-proven acute rejection, all-cause graft loss, and patient death or follow-up loss.

The primary outcome was the evaluation of SNP impact on the risk of developing NODAT within the first year after transplantation. Based on the definition of the American Diabetes Association, NODAT was diagnosed when fasting blood sugar was higher than 126 mg/dL six months after transplantation, or when insulin or oral hypoglycemic agents were required for treatment [23]. The control group consisted of renal transplant recipients who did not meet NODAT criteria during the follow-up period.

#### *2.4. Statistical Analysis*

Continuous variables were presented as the mean ± standard deviation. Allelic frequencies were analyzed using a chi-squared test between the two groups. Student's t-tests and chi-squared tests were used to evaluate between-group differences for continuous and categorical variables, respectively. For all SNPs, minor allele frequency (MAF), compliance with Hardy-Weinberg equilibrium (HWE), linkage disequilibrium analysis, and the association between rs2237892 and NODAT in different genetic models were assessed using SNPstats software (https://www.snpstats.net/start.htm). A multivariate logistic regression model was used to investigate the confounding effects of clinical variables significantly associated with NODAT and SNP associations. We included clinical covariates according to their weights in univariate testing, and we included clinically fundamental parameters. The confounders used in this analysis were recipient age, recipient sex, BMI, HLA mismatch number, desensitization in HLA incompatibility, ABO incompatibility, use of tacrolimus, use of steroids, biopsy-proven acute rejection, donor age, and deceased donor. Bonferroni correction was used in the association analysis when multiple comparisons were performed. We used multiple inheritance models, including codominant (major allele homozygotes vs. heterozygotes vs. minor allele homozygotes), dominant (major allele homozygotes vs. minor allele homozygotes plus heterozygotes), recessive (major allele homozygotes plus heterozygotes vs. minor allele homozygotes), and log-additive (major allele homozygotes vs. heterozygotes vs. minor allele homozygotes) models. Statistical analyses were performed using SPSS for Windows software (version 20.0; SPSS, Chicago, IL, USA). The significance level was set at *p* < 0.05.

#### **3. Results**

#### *3.1. Baseline Clinical Characteristics and SNP Information*

The incidence of NODAT in this study population was 23.0% (254/1102 patients). Baseline characteristics of recipients are summarized in Table 1. Transplant recipients who developed NODAT were significantly older, tended to be male, and had higher BMI scores than those who did not develop NODAT. Donor age in the NODAT group was significantly higher than in the control group. Desensitization treatment for HLA incompatibility was used more frequently in the control group. There was no difference between the two groups in the incidence of biopsy-proven acute rejection, or the use of tacrolimus or steroids as maintenance immunosuppressant treatments.

We excluded *AGMAT* rs11580170 from further analysis because it was in strong linkage disequilibrium with *DNAJC16* rs7533125 (*r*<sup>2</sup> = 0.99). Of rs1494558 and rs2172749 in *IL7R*, only rs2172749 was analyzed, because these SNPs were also in linkage disequilibrium (*r*<sup>2</sup> = 0.98). Of the 15 SNPs tested, 14 were consistent with HWE (*p* > 0.05). While *DNAJC16* rs7533125 violated HWE in the control group (*p* = 0.037), minor allele frequency (MAF) did not deviate from that of the East

Asian population [24]. Therefore, we included *DNAJC16* rs7533125 in the genetic association test. We additionally excluded *TCF7L2* rs7903146 and *NPPA* rs198372 in the association test, because MAF was less than 0.05 (frequency of T allele at *TCF7L2* rs7903146, 0.02; and frequency of A allele at *NPPA* rs198372, 0.01).


**Table 1.** Baseline demographics and characteristics of the study population.

BMI = body mass index; and NODAT = new-onset diabetes after renal transplantation.

#### *3.2. Allelic Frequency and Association between SNPs and NODAT*

The allele frequencies of the genetic polymorphisms in the NODAT and control groups are summarized in Table 2. The allelic frequency of the T allele at *KCNQ1* rs2237892 was significantly lower in patients with NODAT compared to that in the control group (0.30 vs. 0.39; *p* = 8.5 <sup>×</sup> 10−5). The C allele at *CDKAL1* rs10946398 had a higher frequency in the NODAT group, with marginal statistical significance (0.52 vs. 0.47; *p* = 0.080).

We examined the genetic association between SNPs and NODAT in an allele-specific pattern (Table 3). Univariate analyses showed that the T allele at *KCNQ1* rs2237892 was significantly associated with decreased risk of NODAT (odds ratio (OR) 0.66, 95% confidence interval (CI) 0.53–0.82; *p* = 1.3 <sup>×</sup> 10−4). The C allele at *CDKAL1* rs10946398 was associated with a 1.2-fold higher risk for development of NODAT (95% CI 0.98–1.46; *p* = 0.078). However, none of the other SNPs evaluated in this study (*ATP5F1P6* rs10484821, *DNAJC16* rs7533125, *CELA2B* rs2861484, *CASP9* rs2020902, *NOX4* rs1836882, *INPP5A* rs4394754, *IL7R* rs2172749, *IL17R* rs4819554, *IL17RB* rs1025689, *IL17RB* rs1043261, and *PLXDC1* rs72823322) were significantly associated with NODAT. The association between *KCNQ1* rs2237892 and NODAT was enhanced when evaluated using multivariate logistic regression analysis (OR 0.63, 95% CI 0.51–0.79; *<sup>p</sup>* <sup>=</sup> 5.5 <sup>×</sup> <sup>10</sup><sup>−</sup>5). However, no other SNPs were significantly associated with NODAT in the multivariate logistic regression analysis.


**Table 2.** Allele frequencies of polymorphisms previously associated with NODAT.

NODAT = new onset diabetes after renal transplantation; Chr = chromosome; MAF = minor allele frequency; and SNP = single nucleotide polymorphism.

**Table 3.** Allele-based incidence and risk of NODAT.


NODAT = new onset diabetes after renal transplantation; CI = 95% confidence interval; OR = odds ratio; and SNP = single nucleotide polymorphism. \* Adjusted for recipient age, recipient sex, BMI, HLA mismatch number, desensitization in HLA incompatibility, ABO incompatibility, use of tacrolimus, use of steroids, biopsy-proven acute rejection, donor age, and deceased donor.

#### *3.3. Genotype Distribution and Association between KCNQ1 rs2237892 and NODAT*

We tested the effect of *KCNQ1* rs2237892 genotype on NODAT using a multiple inheritance model as shown in Table 4). In the codominant model, the TT genotype at rs2237892 was associated with the lowest risk for development of NODAT, compared to the CC genotype (OR 0.41, 95% CI 0.25–0.67; *<sup>p</sup>* <sup>=</sup> 4.7 <sup>×</sup> <sup>10</sup><sup>−</sup>4). In the dominant model, the CT/TT genotype was also associated with a reduced risk for development of NODAT (OR 0.56, 95% CI 0.42–0.76; *<sup>p</sup>* <sup>=</sup> 2.0 <sup>×</sup> <sup>10</sup><sup>−</sup>4). The T allele significantly reduced the risk of NODAT compared to the CC genotype in the log-additive model. However, no significant differences were observed in the recessive model with Bonferroni correction.


**Table 4.** NODAT incidence and risk of *KCNQ1* rs2237892 in multiple inheritance models.

NODAT = new onset diabetes after renal transplantation; CI = 95% confidence interval; and OR = odds ratio. \*Adjusted for recipient age, recipient sex, BMI, HLA mismatch number, desensitization in HLA incompatibility, ABO incompatibility, use of tacrolimus, use of steroids, biopsy-proven acute rejection, donor age, and deceased donor.

#### **4. Discussion**

In the present study, using samples from a large cohort of renal transplant recipients, we examined the association of 13 SNP pairs and candidate genes for risk of NODAT development. Of the studied variants, there was a significant difference in the frequency of the T allele at *KCNQ1* rs2237892 between the NODAT and control groups, and this allele showed an independent association with NODAT. The TT and CT genotypes of *KCNQ1* rs2237892 were associated with a significantly reduced risk for development of NODAT in codominant, dominant, and log-additive models. These findings suggested that the genetic variant of *KCNQ1* is a significant contributor to the development of NODAT in renal transplant recipients.

Although NODAT results from the combined effect of insulin resistance and β-cell dysfunction, several recent studies have shown that β-cell dysfunction is the main contributing factor for the development of NODAT [3,25,26]. *KCNQ1* rs2237892 and *CDKAL1* rs10946398 were identified as a susceptibility gene for type 2 diabetes in GWAS, and each of these genes is associated with β-cell dysfunction [27–31]. Previous studies with type 2 diabetes risk genes suggested an association between *KCNQ1* rs2237892 and NODAT [19]. Our study also validated that variant rs2237892 of the T allele was associated with decreased risk for development of NODAT compared to the C allele. Similarly, *CDKAL1* rs10946398 was also associated with NODAT, as reported in a study that used a candidate gene approach in patients who underwent transplantation [18,19]. However, our data did not confirm this association. These findings suggested that *KCNQ1* is a more robust and influential indicator of β-cell dysfunction in renal transplant recipients.

*KCNQ1* encodes a subunit of the voltage-gated K<sup>+</sup> channel, which is expressed in pancreatic islets [32]. In the *KCNQ1*-overexpressing pancreatic β-cell line, the density of the K+ current increased significantly and affected the pancreatic cell membrane action potential [33]. Therefore, *KCNQ1* overexpression contributes to impairment of glucose-stimulated insulin secretion, and a specific *KCNQ1* blocker also stimulates insulin secretion [34]. In addition, allelic mutation of *KCNQ1* results in up-regulation of the neighboring gene, cyclin-dependent kinase inhibitor 1C, which encodes a cell cycle inhibitor and leads to reduction in pancreatic β-cell mass [35]. Therefore, we suggest that variant *KCNQ1* induces impaired β-cell function and reduced β-cell mass, and this biological function could be a potential underlying mechanism for the association between *KCNQ1* variants and increased risk for NODAT development.

Three types of *KCNQ1* SNPs were evaluated as potential risk factors for the development of NODAT in Spanish patients who received kidney transplants from deceased donors [17]. *KCNQ1* rs2237895, rs2237892, and rs8234 were genotyped, and SNP rs2237895, but not rs2237892, was found to be associated with an increased risk for development of NODAT in the first year after transplantation. This apparent discrepancy could be due to the allele frequencies of these SNPs. The T allele frequency at rs2237892 was reported to be 0.34–0.36 in the East Asian population, but only 0.04–0.08 in the European

population [36]. Consequently, lower MAF at rs2237892 was not significantly associated with NODAT in Spanish transplant recipients. Therefore, we suggest that different genetic backgrounds should be considered when attempting to determine the risk of development of NODAT using *KCNQ1* genetic variants as indicators.

In a recent GWAS, numerous variants were found to be associated with risk for the development of NODAT [21]. *ATP5F1P6*, *CELA2B*, *CASP9*, *NOX4*, and *INPP5A* were identified as risk genes in Caucasian renal transplant recipients. These genetic variants were implicated in β-cell apoptotic pathways, but not insulin resistance, suggesting that β-cell apoptosis was a critical component of NODAT pathogenesis. However, our study did not find a significant association between NODAT and any SNPs from this GWAS. Three possible factors might explain this inconsistency: First, the β-cell apoptotic pathways could be a weak contributor to the development of NODAT in Asian compared to Caucasian recipients of a renal transplant. Second, a different definition of NODAT phenotype might have resulted in dissimilar findings in the two studies. Third, the limited sample size in the GWAS may have less power to detect significant associations [37].

Inflammatory cytokines are involved in insulin action and insulin secretion. An SNP within the gene encoding the IL-7R chain was found to be associated with type 1 diabetes mellitus [38,39]. Moreover, our previous study showed that genetic variants of *IL-7R*, *IL-17R*, and *IL-17RB* were associated significantly with NODAT [14]. However, the relevant SNPs of these interleukin genes were not associated with NODAT in the exploratory GWAS analysis or the secondary verification analysis [17]. Furthermore, our validation study also showed no meaningful differences in allele frequencies. These findings suggested that the effects of interleukin gene polymorphisms on the risk for development of NODAT were inconclusive, and further studies are necessary to obtain precise results.

The present study had a few limitations. As data regarding family history of type 2 diabetes were not available, the association between family history and the development of NODAT could not be evaluated. In addition, the effect of BMI and weight gain after transplantation was not included in our analysis. Finally, we did not perform an oral glucose tolerance test or HbA1c estimation before kidney transplantation. Therefore, patients with prediabetes might have been included in our study.

In conclusion, our validation study showed a significant association between *KCNQ1* rs2237892 and development of NODAT in a large cohort. Our results suggest that *KCNQ1* might play a crucial role in the pathogenesis of NODAT following renal transplantation. *KCNQ1* variants might be a useful tool to predict NODAT development in renal transplant recipients, and help screen for patients at a higher risk for NODAT.

**Author Contributions:** Conceptualization, S.H.L.; methodology, H.S.H.; software, K.-W.H.; validation, all authors; formal analysis, H.S.H.; investigation, J.S.K., Y.G.K., J.Y.M., K.H.J.; resources, J.S.K., Y.G.K., J.Y.M., K.H.J.; data curation, H.S.H.; writing—original draft preparation, H.S.H.; writing—review and editing, S.H.L.; visualization, H.S.H.; supervision, S.H.L.; project administration, J.S.K., Y.G.K., J.Y.M., K.H.J., S.H.L.; funding acquisition, S.H.L.

**Funding:** This research was funded by the Korean Health Technology R&D Project, Ministry of Health & Welfare, Republic of Korea (grant no. HI13C1232), and Research of Korea Centers for Disease Control and Prevention (2014-ER6301-00, 2014-ER6301-01, 2014-ER6301-02, 2017-ER6301-00, 2017-ER6301-01).

**Acknowledgments:** The authors appreciate the support and cooperation of Korean Organ Transplantation Registry Study Group: Jin Min Kong 1, Oh Jung Kwon 2, Myung-Gyu Kim 3, Sung Hoon Kim 4, Yeong Hoon Kim 5, Joong Kyung Kim 6, Chan-Duck Kim 7, Ji Won Min 8, Sung Kwang Park9 , Yeon Ho Park 10, Inwhee Park 11, Park Jae Berm 12, Jung Hwan Park 13, Jong-Won Park 14, Tae Hyun Ban 15, Sang Heon Song 16, Seung Hwan Song 17, Ho Sik Shin 18, Chul Woo Yang 19, Hye Eun Yoon 20, Kang Wook Lee 21, Dong Ryeol Lee 22, Dong Won Lee 23, Sam Yeol Lee 24, Sang-Ho Lee 25, Jung Jun Lee 26, Lee Jung Pyo 27, Jeong-Hoon Lee 28, Jin Seok Jeon 29, Heungman Jun 30, Kyung Hwang Jeong 31, Ku Yong Chung 32, Hong Rae Cho 33, Ju Man Ki 34, Dong-Wan Chae 35, Soo Jin Na Choi 36, Duck Jong Han 37, Seungyeup Han 38, Kyu Ha Huh 39, Jaeseok Yang 40, Curie Ahn 41; <sup>1</sup> Department of Nephrology, BHS Hanseo Hospital, <sup>2</sup> Department of Surgery, College of Medicine, Han Yang University, <sup>3</sup> Department of Internal Medicine, Korea University Anam Hospital, <sup>4</sup> Department of Surgery, Yonsei University Wonju College of Medicine, Wonju Severance Christian Hospital, <sup>5</sup> Department of Internal Medicine, Inje University Busan Paik Hospital, <sup>6</sup> Department of Internal Medicine, Bongseng Memorial Hospital, <sup>7</sup> Department of Internal Medicine, School of Medicine, Kyungpook National University

Hospital, <sup>8</sup> Division of Nephrology, Department of Internal Medicine, Bucheon St. Mary's Hospital, <sup>9</sup> Department of Internal Medicine, Chonbuk National University Medical School, <sup>10</sup> Department of Surgery, Gil Medical Center, Gachon University College of Medicine, <sup>11</sup> Department of Nephrology, Ajou University School of Medicine, <sup>12</sup> Department of Surgery, Samsung Medical Center, <sup>13</sup> Konkuk University School of Medicine, Department of Nephrology, <sup>14</sup> Department of Nephrology, Yeungnam University Hospital, <sup>15</sup> Division of Nephrology, Department of Internal Medicine, Eunpyeong St. Mary's hospital, <sup>16</sup> Organ Transplantation Center and Department of Internal Medicine, Pusan National University Hospital, <sup>17</sup> Department of Surgery, Ewha Womans University Medical Center, <sup>18</sup> Kosin University College of Medicine, Department of Internal Medicine, Division of Nephrology, <sup>19</sup> Division of Nephrology, Department of Internal Medicine, Seoul St. Mary's hospital, <sup>20</sup> Department of Internal Medicine, Incheon St. Mary's Hospital, <sup>21</sup> Department of Nephrology, Chungnam National University Hospital, <sup>22</sup> Division of Nephrology, Department of Internal Medicine, Maryknoll Medical Center, <sup>23</sup> Division of Nephrology, Department of Internal Medicine, Pusan National University School of Medicine, <sup>24</sup> Department of Surgery, Kangdong Sacred Heart Hospital, <sup>25</sup> Department of Nephrology, Kyung Hee University Hospital at Gangdong, <sup>26</sup> Department of Surgery, CHA Bundang Medical Center, <sup>27</sup> Department of Nephrology, SNU Boramae Medical Center, <sup>28</sup> Department of Surgery, Myongji Hospital, <sup>29</sup> Department of Internal Medicine, Soonchunhyang University Seoul Hospital, <sup>30</sup> Department of Surgery, Inje University Ilsan Paik Hospital, <sup>31</sup> Department of Internal Medicine, Kyung Hee University College of Medicine, <sup>32</sup> Department of Surgery, Ewha Womans University Mokdong Hospital, <sup>33</sup> Department of Surgery, Ulsan University Hospital, <sup>34</sup> Department of Surgery, Gangnam Severance Hospital, Yonsei University College of Medicine, <sup>35</sup> Division of Nephrology, Seoul National University Bundang Hospital, <sup>36</sup> Department of Surgery, Chonnam National University Medical School, <sup>37</sup> Department of Surgery, Asan Medical Center, <sup>38</sup> Department of Internal Medicine, Keimyung University School of Medicine, <sup>39</sup> Department of Transplantation Surgery, Severance Hospital, <sup>40</sup> Department of Surgery, Seoul National University Hospital, <sup>41</sup> Department of Nephrology, Seoul National University Hospital.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
