Relation of Wnt Signaling Pathway Inhibitors (Sclerostin and Dickkopf-1) to Left Ventricular Mass Index in Maintenance Hemodialysis Patients

Background Left ventricular hypertrophy (LVH) is common in hemodialysis (HD) patients. It predicts poor prognosis. Several inhibitors regulate Wnt canonical pathways like Dickkopf-related protein-1 (Dkk-1) and sclerostin. Objectives To investigate the relationship between serum sclerostin, Dkk-1, left ventricular mass (LVM), and LVM index (LVMI) in HD patients. Methods This is a cross-sectional study including 65 HD patients in our HD unit. Patients were divided into two groups according to LVMI (group 1 with LVMI < 125 gm/m2 (N = 29) and group 2 with LVMI > 125 gm/m2 (N = 36)). Echocardiographic evaluation of the LVM, aortic, and mitral valves calcification (AVC and MVC) was done. Serum levels of sclerostin and Dkk-1 and patients' clinical and biochemical data were recorded. Results Group 2 showed significantly higher age, blood pressure, AVC, and MVC and significantly lower hemoglobin, sclerostin, and Dkk-1 levels. LVM and LVMI had a significant linear negative correlation to both serum sclerostin and Dkk-1 (r = −0.329 and −0.257, P=0.01 and 0.046 for LVM; r = −0.427 and −0.324, P=0.001 and 0.012 for LVMI, resp.). Serum Dkk-1 was an independent negative indicator for LVM and LVMI in multiple regression analyses (P=0.003 and 0.041 with 95% CI = −0.963 to −0.204 and −0.478 to −0.010, resp.). Conclusion Serum sclerostin and Dkk-1 were significantly lower in HD patients with increased LVMI > 125 gm/m2, and both had a significant linear negative correlation with LVM and LVMI. Dkk-1 was a significant negative independent indicator for LVM and LVMI in HD patients.


Introduction
Left ventricular hypertrophy (LVH) is common structural remodeling in patients with end-stage renal disease, and its presence predicts a poor prognosis [1]. Echocardiographic diagnosis of LVH is based on cutoff values developed from formula formed from population-based studies which indexed the left ventricular mass (LVM) to the body surface area (BSA) [2].
Wnt/β-catenin signaling pathway is an essential positive regulated signaling network for cardiovascular diseases [3]. β-Catenin increases the expression of target genes associated with cell adhesion and involves the regulation of angiogenesis and atherosclerosis [4].
In addition, β-catenin plays a crucial role in heart failure caused by afterload-induced cardiac hypertrophy [5]. It interacts with the transforming growth factor-β (TGF-β) signaling pathway to exacerbate cardiac fibrosis and aggravate chronic heart failure [6].
Several inhibitors regulate the Wnt canonical pathway, among them Dickkopf-related protein-1 (Dkk-1) and sclerostin (Scl) [7]. Sclerostin is a 190-residue glycoprotein, which is expected to contain a cysteine-knot motif and belongs to the DAN/Cerberus family of proteins [8]. e secreted glycoprotein Dkk-1, a member of the Dickkopf family, is known to antagonize Wnt/β-signaling by interaction with the LRP5/6 (low-density lipoprotein receptor-related protein 5/6).
e Wnt/β-catenin signaling pathway is an important mediator in cardiovascular disease and influences inflammation and vascular calcification [9].
Serum sclerostin levels significantly vary with age and are higher in male patients with stage 3b and 4 CKD than in females with the same stages [10]. Previous studies showed that the expression of Wnt/β-catenin signaling inhibitors such as sclerostin and Dkk-1 may attenuate further vascular calcification [11,12].
We aimed to define the relationship between serum sclerostin, Dkk-1, left ventricular mass (LVM), and LVM index (LVM/BSA) in maintenance hemodialysis (HD) patients.

Study Design.
is is a cross-sectional study involving patients on regular HD in MNDU during the period Jan 2016-Jan 2017. 65 patients were recruited. e patients were grouped according to the presence or absence of left ventricular hypertrophy (LVH) calculated by LVM/BSA (LVMI) into two groups, group 1 with LVMI < 125 gm/m 2 and group 2 with LVMI > 125 gm/m 2 , and the cut point that indicated LV hypertrophy was 125 gm/m 2 [13]. All patients were undergoing HD three times/week, 4 h/session using bicarbonate-based dialysate with calcium concentrations of 1.5 mmol/L. Exclusion criteria were age <18 or >75 years, HD duration less than 6 months, patients who had a rheumatic valvular disease or underwent prosthetic valve replacement, and cardiomyopathic patients with EF < 40%. All patients were subjected to history taking with stress on age, smoking state, HD duration, cause of ESRD, drug history and current treatment, associated comorbidities, diabetes mellitus, hypertension, chronic liver disease, CVD, surgical history especially previous renal transplantation, and parathyroidectomy. Estimated glomerular filtration rates (eGFRs) were calculated using the Modification of Diet in Renal Disease (MDRD) equation [14]. Physical examination including weight, height, waist, midarm circumference, blood pressure, edema of lower limb, chest, heart, and abdomin, and AV fistula examination was done.

Radiology Assessment
Echocardiography. An expert echocardiographer, who was unaware of the patients' data, performed echocardiographic measurements according to the recommendations of the American Society of Echocardiography [16]. Patients were examined prior to the HD session lying in a left lateral position in a semidark room; a two-dimensional assessment of the aortic valve and mitral valve was done using Medison SonoAce X6 device. Scoring of mitral calcification was done according to the Wilkins calcification [17] and grading of the aortic valve was done according to a previous study by Tenenbaum et al. [18].

Calculation of LV Myocardial Performance Index (MPI).
MPI was calculated using (ICT + IRT)/ET formula. e mean normal value of the Tei index is 0.39 ± 0.05 for the LV [20]. Mitral inflow was recorded by conventional pulsed Doppler and tissue Doppler to reveal the left ventricular diastolic function by measuring E (early diastolic) velocity, A (late diastolic) flow velocity, also E 1 (early diastolic) annular velocity, and A 1 (late diastolic) annular velocity and calculate E/A ratio and E 1 /A 1 ratio, respectively.

Statistical Analysis.
e normality of data was first tested with the Shapiro-Wilk test. Qualitative data were presented by frequency tables (frequency and percentages). Quantitative variables were presented by central indices (mean ± standard deviation) for normally distributed variables and median (minimum-maximum) for nonnormally distributed variables. Pearson's correlation was used to correlate continuous normally distributed data while Spearman's correlation was used to correlate ordinal and nonnormally distributed data. We compared nonnormally distributed and ordinal variables between qualitative groups using the Mann-Whitney U test and Kruskal-Wallis H test. All statistical analyses were performed using SPSS version 24 (IBM Corp., Armonk, NY, USA). A P value ≤ 0.05 was considered statistically significant. ). Regarding smoking, current smokers were 6 (18.8%), ex-smokers were 6 (18.8%), and nonsmokers were 20 (62.4%).

Discussion
e accumulated data allow us to consider the disturbances in the FGF-23-Klotho-sclerostin ratio as one of the early markers of CKD advancement, disorders of mineral metabolism developing, and cardiovascular prognosis [21]. Alteration in the ratio of FGF-23, serum Klotho, and serum sclerostin can be regarded as an independent early marker of cardiovascular morbidity and overall prognosis of patients with CKD [22]. 55.6% of our patients were men with a mean age of 46.1 ± 16.2 years. e majority of them were hypertensive, while diabetes and CAD were present in 13% and 10% of them, respectively. HB concentration was 9.38 ± 1.65 mg/dl. Serum albumin, calcium, and phosphorus were within normal ranges in most patients.
With regard to the comparative analysis of the studied groups, we found a significant increase in age, AVC, and MVC in the second subgroup (with LVM/BSA > 125 gm). A significant increase in systolic and diastolic blood pressures was noted in subgroup 2 despite the nonsignificant difference in the percentage of hypertensive patients in the studied groups which may be attributed to better control of blood pressure and adherence to treatment in group 1 and it also indicates that the second group was subjected to a greater afterload burden. ere was a nonsignificant decrease in (E/ A ratio) by conventional Doppler study in the subgroup 2 versus subgroup 1 (P � 0.069; Table 2) denoting more diastolic dysfunction and more increase in the preload in subgroup 2 patients. When tissue Doppler was used, the E1/ A1 ratio decreased in group 1 denoting that the tissue Doppler study revealed more patients with diastolic dysfunction in subgroup 1 patients, yet statistically nonsignificant. None of the studied patients in our study had AF.
ere was a significant increase in hemoglobin concentration in subgroup 1 versus subgroup 2, denoting more malnutrition in subgroup 2.
One possible explanation is the start of hemodialysis as it has been reported that, in patients on hemodialysis, the high sclerostin level was negatively correlated with the risk of cardiovascular death and Ji et al. did not include such patients. Chen et al. followed up 84 hemodialysis patients for (12-60 months) and concluded that patients with low baseline serum sclerostin undergoing MHD showed better survival and less CIMT [24]. Zou et al. found no correlation between sclerostin and CVEs in MHD patients [25]. ese contradictory results may be attributed to different ethnic groups, different follow-up period, different assessment and measurement of sclerostin, and different primary endpoints. It is known that PTH suppresses the expression of sclerostin and/or Dkk-1 decreasing their levels [26] in contradiction to our results where it displayed a nonsignificant positive  is could be due to increased extraosseous production of sclerostin and Dkk-1 and hyperphosphatemia, increased calcitonin exposure, and absent renal clearance of both molecules [27,28].
Both molecules did not correlate significantly to different treatments of HPT in our study (calcium supplement, active vitamin D and analogs, cinacalcet, sevelamer HCl, and parathyroidectomy) in contradiction to Kuczera et al., 2016, who reported an increase in sclerostin level (in 42 patients out of 58 included in their study) after cinacalcet therapy and related that to the decrease in PTH levels, but actually, the other 16 patients whose PTH levels did not decrease showed an increase in their sclerostin level over their study period [29] which means that further larger RCT are needed to clarify this dilemma.
In our study, sclerostin and Dkk-1 did not correlate significantly to Kt/v, calcium, phosphorus, ALP, iron, TSAT, TIBC, and ferritin. Both molecules correlated negatively to EPO treatment but only Dkk-1 showed significance (r � −0.372, P � 0.003). is finding is overlooked and rarely found in the literature. Sharba and Al Zahid, 2016, found a positive correlation between EPO and sclerostin in their study on 21 male patients (12 on dialysis and 9 not on dialysis; duration of dialysis 20.9 ± 3.25 months) [30]. More detailed studies are needed to explore this interesting issue.
A study done by Stróżecki and colleagues concluded that patients with vascular calcification (VC) were older and had higher LVMI and no significant differences were found with respect to PTH, phosphorus calcium which is in accordance with our results. VC coexists with hypertrophy of the left ventricle especially when both valves are calcified. Even shortlived incidents involving increased product Ca x P can lead to cardiovascular calcification [31,32].
LVH has a prevalence of approximately 40% in patients with chronic kidney disease (CKD), and it progressively increases with CKD progression up to 75% in ESRD patients [33,34]. Foley et al. followed 596 patients on hemodialysis with no previous history of heart disease to determine whether the frequency of LVH correlates with the length of the dialysis. After 18 months of dialysis, the author recorded that 62% of patients had an increased LV mass index and 49% had developed LV failure [35]. CKD patients are faced with both pressure and volume overload states; however, sustained overload in combination with CKD-associated factors such as SHPT, RAAS activation, and anemia may result in maladaptive LVH until the final picture of uremic cardiomyopathy ensues [36][37][38][39].
Our results reveal that there was a significant negative correlation of sclerostin and Dkk-1 with left ventricular mass (LVM). LVM had a significant negative correlation with HDL, while it showed a significant positive correlation with age and a significant positive correlation with BSA, hypertension, AVC, MVC, and dry weight.
With linear regression analysis for LVM as a dependent factor, there was a nonsignificant prediction of age, serum sclerostin, aortic valve calcification, and mitral valve calcification for LVM; on the other hand, hypertension had a significant positive independent prediction for LVM, while serum HDL and serum DDK-1 were negative significant independent indicators for LVM.
As regards correlations with LVM/BSA (LVMI), there was a significant negative correlation of serum sclerostin and Dkk-1 with LVMI and a significant negative correlation with hemoglobin, albumin, midarm circumference, and dry weight with LVMI, while there was a highly significant positive correlation with age, AVC, and MVC with LVMI. With linear regression analysis for LVMI, there was a nonsignificant prediction of age, HDL, MVC, and serum sclerostin for LVMI; on the other hand, serum DDK-1 was a significant negative indicator for LVMI and hypertension was a weak significant positive predictor for LVMI (P value � 0.05).
Our results are in disagreement with Brovko and colleagues, 2018, who had done a study on 131 CKD Russian patients (stages 1-5, average age 42.4 ± 13.7 years) and revealed that, with univariate analysis, serum sclerostin levels correlated positively with LVM index (r � 0.545; P < 0.01) but with multivariate regression, it was negatively correlated to CVC and suggested that it may protect the heart and vessels against calcification with CKD advancement [40].
It is well known that Wnt-β-catenin signaling pathway inhibition by sclerostin and Dkk-1 protects against cardiac AV and MV calcifications and both have a significant negative correlation with both valves' calcification [41,42]. Our study revealed that AVC and MVC had a significant negative correlation with sclerostin but a nonsignificant negative correlation with Dkk-1 which coincides with Yang et al., 2015, who reported that circulating sclerostin but not Dkk-1 is inversely International Journal of Nephrology associated with aortic calcifications and future cardiovascular events [43]. Krishna et al. reported that sclerostin decreases the expression of genes involved in matrix degradation and calcification and thereby inhibits atherosclerosis; it is likely that sclerostin could function as an inhibitor of vascular calcification [44].
On the contrary, other studies showed that elevated serum sclerostin levels were seen in patients with aortic valve calcification with increased upregulation of sclerostin mRNA [45]. Several elements could have contributed to these conflicting results. Studies investigated the presence of vascular calcification at different anatomical locations and used different statistical analyses (by adjusting for different potential confounding factors). In addition, differences in the time period between administration of enoxaparin (or other low molecular weight heparins that are used as anticoagulants) and blood collection could also be important, since enoxaparin stimulates the release of sclerostin into the circulation. is also implies that the heterogeneity of study populations-patients at different CKD stages, with or without dialysis treatment, whether or not they receive low molecular weight heparins-contributes to the observed inconsistency. Lastly, it is known that there are large discrepancies between sclerostin assays. e antibodies that are used in the distinct assays bind different epitopes; therefore, some antibodies will capture only the intact sclerostin molecule, while others might also bind to sclerostin fragments [46]. In our study, sclerostin and Dkk-1 had a significant negative correlation with LVM and LVM/BSA. Moreover, Dkk-1 appeared to be an independent indicator of LVM and LVM/BSA and they may be acting as protectors against LV hypertrophy or may be only associated with LV hypertrophy in CKD patients.
Van de Schans and colleagues demonstrated that interruption of Wnt signaling in the mice lacking the Dvl-1 gene delays the onset of pressure overload-induced cardiac hypertrophy. erefore, the Wnt/Frizzled pathway may provide novel therapeutic targets for antihypertrophic therapy [47].
Because Wnt signaling is very complex and its effects may be quite varied according to the cell system, specific ligands/ receptors involved, and timing of any interventions which presents a challenge to investigators, but it offers rich possibilities for new insights into cardiac pathophysiology and the identification of new therapeutic targets [48].
Limitations of this study include the relatively small population studied in a single-center and the cross-sectional design. However, to our knowledge, there are only

Data Availability
All data analyzed during this study are included within this manuscript.
Ethical Approval e study was approved by the Institutional Research Board (IRB) of the Faculty of Medicine, Mansoura University, in accordance with the Declaration of Helsinki.

Consent
All participants signed written informed consent before the start of the study.

Conflicts of Interest
All the authors declare no conflicts of interest in this work.  International Journal of Nephrology 9