Patients with Chronic Kidney Disease (CKD) are upto 3.5 times more likely to die from diseases of heart and blood vessels (Cardiovascular Disease-CVD). Vitamin D insufficiency is very common in CKD and associated with CVD. Animal studies have shown an improvement in heart size and function with Vitamin D therapy, although evidence in humans is lacking. The proposed study will test if oral Vitamin D treatment, in deficient CKD patients, will improve heart enlargement and function. With these proposed changes the investigators expect to reduce CVD and deaths in patients with CKD.

Impact of Vitamin D supplementation on Cardiac Hypertrophy and Function in Chronic Kidney Disease Patients- A randomised placebo controlled trial.

1. Lay summary:

   Patients with Chronic Kidney Disease (CKD) are up to 3.5 times more likely to die from diseases of heart and blood vessels (Cardiovascular Disease-CVD). Vitamin D insufficiency is very common in CKD and associated with CVD. Animal studies have shown an improvement in heart size and function with Vitamin D therapy, although evidence in humans is lacking. The proposed study will test if oral Vitamin D treatment, in deficient CKD patients, will improve heart enlargement and function. With these proposed changes the investigators expect to reduce CVD and deaths in patients with CKD.

2. Abstract of proposed investigation:

Traditional risk factors inadequately explain the disproportionate increase in CV risk in CKD. Hypovitaminosis D is common in CKD and associated with increased CV mortality. In animal experiments, vitamin D reduces LV mass (LVM) and improves systolic function. In dialysis patients, observational studies show that vitamin D therapy reduces CV mortality and improves diastolic function. However, benefit of native vitamin D repletion on cardiac structure and function has not been investigated in humans.

The investigators propose to test in a randomised, placebo-controlled trial, if 40 weeks of oral Colecalciferol, in predialysis CKD patients, will result in reduction of LV mass and improvement in systolic/diastolic function along with cardiac extracellular matrix composition.

Currently, NICE does not recommend routine measurement of vitamin D in CKD patients. If effective in this setting, further outcome trials will prove if this inexpensive and easily administered therapy can be routinely prescribed to prevent morbidity and mortality.

3. Background to the project and pilot data: (i) CVD in CKD: CV death rates are up to 3.4 times higher in patients with CKD compared to the general population(1). The investigators and others have demonstrated a graded relationship between CV events and death with declining renal function(2). Traditional risk factors only partially explain the elevated risk. This highlights the importance of non-traditional risk factors such as inflammation, oxidative stress, and vitamin D deficiency as modifiable CV risk factors in CKD.

(ii) Vitamin D deficiency in the general population and in CKD Hypovitaminosis D is common in general population. A study by Thadhani's group of 15,088 adults has shown the prevalence of Vitamin D insufficiency to be between 40-80% in the US general population(3). CKD patients are at an increased risk of hypovitaminosis D. In a group of 145 non-dialysis CKD patients from south London, have seen a 85% prevalence of Vitamin D insufficiency [25 hydroxy vitamin D (25(OH)D)< 75 nmol/L], including 32% deficiency[25(OH)D < 37.5nmol/L], with a significantly higher risk among Afro-Caribbeans' and south-Asians'(4).

(iii) Hypovitaminosis D and CV disease: Cross sectional studies show an independent association between low 25 (OH) D and CV mortality(5;6). In the dialysis population, observational studies by Thadhani et al show an increase in early CV mortality in patients with low 25 (OH) D levels (OR=1.9). Vitamin D supplementation is associated with better survival on dialysis (OR=0.6)(7).

(iv) Vitamin D improves cardiac function and morphology: The Vitamin D endocrine system is ubiquitous in human tissues, including the myocardium and the vascular endothelium. There is now mounting evidence from animal studies that Vitamin D directly affects cardiac function and structure through a genomic action on cardiac myocytes. Vitamin D knockout mice show an increase in cardiomyocyte hypertrophy, cardiac fibrosis and heart weight (8). More importantly, treatment with vitamin D in spontaneously hypertensive heart failure rats and wild type mice has been shown to attenuate cardiac hypertrophy and fibrosis, and modulates cardiac myocyte contractility, respectively (9;10).

In an animal model of Dahl salt sensitive rats treated with active vitamin D for 6 weeks, Thadhani's group demonstrated a reduction in heart weight, LVM, posterior wall thickness and increased fractional shortening. There was a significant reduction in plasma BNP and renin levels in rats fed high salt (HS) along with Vitamin D, compared to rats fed HS diet alone. There was no difference in BP between the two groups, thus reaffirming a direct action of vitamin D on cardiac morphology and function, as opposed to a haemodynamic effect. Most importantly, they also demonstrated an improvement in diastolic function, and reduction in septal and posterior wall thickness in a small observational study of haemodialysis patients, treated with active vitamin D, suggesting that the findings from animal models can be replicated in humans(11). Moreover, intravenous Vitamin D administration in a group of haemodialysis patients with secondary hyperparathyroidism was associated with a significant reduction in LVM Index (178 ±73 to 165±61 g/m2) on transthoracic echocardiogram over 15 weeks(12). However, no studies thus far have investigated the effects of native vitamin D repletion on cardiac structure and function, as a randomized, double-blind, placebo-controlled trial in predialysis CKD.

(v) Vitamin D and cardiac fibrosis: Recent evidence using gadolinium-enhanced cardiac magnetic resonance (CMR), suggests that cardiac fibrosis is a predominant feature of uremic cardiomyopathy(13). Animal studies have demonstrated an important role of vitamin D in maintaining extracellular matrix (ECM) composition in the heart. In vitamin D receptor knockout mice model there is marked increase in cardiac fibrosis, related to abnormal expression of collagenases(14).

Studies by Diez's group in hypertensive heart disease subjects have identified a reliable serum biomarker that reflects type 1 collagen synthesis in human heart- Carboxy-terminal propeptide of procollagen type 1 (PICP)(15). Serum PICP along with free Matrix metalloproteinases 1 (MMP-1) and free Tissue inhibitor of metalloproteinases 1 (TIMP 1), as markers of collagen degradation accurately reflect collagen metabolism in the heart(15). PICP is released in a 1:1 ratio during the synthesis of type 1 collagen, from its precursor procollagen type 1. Furthermore, peripheral serum PICP accurately reflect amount of collagen type 1 fibres in the heart(16) and can therefore be used to monitor the effects of treatment on collagen content of the heart(17). However, no studies thus far have shown an effect of vitamin D therapy on changes in ECM composition within the human heart.

(vi) Abnormal LV size with systolic and diastolic dysfunction in pre dialysis CKD patients: In a group of 18 consecutive stable stage 4 and 5 CKD patients, deemed suitable for renal transplantation, the investigators have seen significant increase in LVM, interventricular septal (IVS) diameter and posterior wall diameter with systolic and diastolic dysfunction; using a combination of Transthoracic and Cardiac Tissue Doppler Imaging (TDI) echocardiography, compared to normal controls.

(5) Original Hypothesis The investigators propose that 40 weeks treatment with oral Colecalciferol will improve cardiac structure and function in predialysis CKD patients, as a consequence of improvement in cardiac ECM composition and turnover.

The investigators plan to address this central hypothesis through the following specific aims:

(i) Specific Aim 1: To determine the effect of oral Colecalciferol therapy on left ventricular mass as measured by CMR imaging. The investigators expect a 10g reduction in LVM on CMR in active treatment group compared to controls.

(ii) Specific Aim 2: To determine the effect of oral Colecalciferol therapy on cardiac systolic function by cardiac TDI and diastolic function by transthoracic echocardiography.

Systolic function will be measured as an improvement in average longitudinal peak systolic strain rate (SR) and longitudinal end systolic strain (S) assessed at the base, mid wall and apex of the IVS, lateral, inferior and anterior wall on cardiac TDI. Improvement in diastolic function will be measured by early (peak E) and late (peak A), trans-mitral flow and annulus motion, isovolumetric relaxation time (IVRT) and deceleration time (DT) on transthoracic echocardiogram.

These parameters will be tested in randomised placebo controlled trial involving 25 patients in each arm. 25 hydroxy vitamin D3 [25 (OH) D] has been used in this proposal, as CKD stage 3-4 patients would be expected to have adequate stores of 1 α-hydroxylase enzyme to convert 25 (OH) D to active 1,25 (OH)2 D. The investigators will measure the level of 25 hydroxy colecalciferol to ensure adequate repletion. Colecalciferol has been shown to be safe and effective when given at this dose(18;19). To examine the improvement in ECM composition and turnover the investigators will demonstrate reduction in cardiac collagen content as measured by decrease in serum PICP and Tissue inhibitor of metalloproteinases 1(TIMP1) and an increase in serum Matrix metalloproteinase 1 (MMP1). PICP will be corrected for eGFR and bone alkPO4. MMP1/TIMP1 ratio will be used to measure the collagen degradation. Cardiac function pre and post therapy, will also be assessed by measuring plasma N-Terminal proBNP (NT proBNP). Furthermore, a biobank of whole cells, mRNA and DNA, blood and urine will be created for future analysis of novel genes or biological pathways.

(6) Experimental detail and design of proposed project

(i) Study Population: Vitamin D naive patients will be recruited from the general nephrology outpatient clinics from St. George's, Guy's and Kingston hospitals. Patients with 25 (OH) D deficiency [(25 (OH) D < 37.5 nmol/L] (Kidney Disease Outcomes Quality Initiative guidelines) on ACEi/ARB therapy, will be enrolled. Cardiac recruitment criteria will be left ventricular mass index (LVMI) between 80-160 g/m2 for females and 100-160 g/m2 for males (these figures correspond to the upper tertile of normal LV mass index (LVMI) on TTE based on Framingham Heart study (20).and those with LVMI 1SD above the mean values observed in CKD3-4, based on the CREATE study) (21)

(ii) Patient randomization, blinding and follow-up: A clinical trials pharmacist will oversee the randomisation into two equal groups for comparison, blinding and dispensing the medications.

The stages of the trial are summarised below:

• 50 patients with 25 (OH) D levels < 37.5 nmol/L will be recruited. Patient demographics, anthropometric data and pertinent clinical history would be obtained upon enrollment. Blood will be collected for calcium, phosphate, parathyroid hormone levels, bone alkPO4, and high sensitivity C reactive protein (hsCRP). Biomarkers of cardiac fibrosis (serum PICP), ECM turnover (MMP1 and TIMP1) and cardiac function (plasma NT proBNP) will be measured. Transthoracic Echocardiography will be performed at baseline to fulfill recruitment criteria and diastolic function. Cardiac TDI function and CMR will be performed at baseline for systolic function and LVM, respectively.
• Patients will be randomised into 2 equal groups to receive 4 directly observed oral doses of 150,000 IU in 50 mls vegetable oil colecalciferol (Martindale, UK) at 0, 10, 20 and 30 weeks or matching placebo.
• At 40 weeks blood samples will be collected for repeat haematology, biochemistry and biomarkers of collagen synthesis and metabolism as mentioned above. Changes in LV geometry and function will be re-measured using CMR, Transthoracic and cardiac TDI echocardiography, respectively.

Changes in the pre and post treatment measurements will be compared in the active and placebo group. All patients will continue to be followed up in the general nephrology clinic during the trial and subsequently upon trial completion.

(iv) CMR protocol: All subjects will be imaged on a 1.5T Philips Intera MRI Scanner with dedicated 32 channel cardiac coil. MRI protocol will include a plan scan, SENSE reference scan, interactive scan to identify 4 chamber, 2 chamber long axis and short axis geometries of the heart. Breath hold Steady State Free Procession 4 chamber, 2 chamber and short axis stack (1.5mm, 1.5mm, 10mm) cine (30 cardiac phases) will be acquired. Ventricular mass will be calculated using a viewforum work station (Philips Healthcare) by delineating the diastolic left ventricular endocardial and epicardial borders(figure 2).

(v) Cardiac Tissue Doppler Imaging and Transthoracic Echocardiography: Cardiac TDI will be recorded for longitudinal deformation. Data will be obtained from septum, lateral, anterior and inferior wall using narrow sector (typically 120). Longitudinal S and SR will be calculated over a computation area of 10 mm. SR and S averaged over 3 cardiac cycles will be analysed using dedicated software (SPEQLE, Univ. Leuven) (figure 3). The intra- and inter- observer variability of this approach is less than 10%(22).

Data will be obtained from parasternal and apical views using a GE Vivid 7 scanner. LV volumes will be measured using the Teichhloz formula. Diastolic function will be assessed as previously described by our group(17). Global radial function will be assessed by both endocardial and mid-wall fractional shortening. All measurements will be performed using Echo-Pac (GE) workstation and averaged over 3 individual measures.

(vi) Analysis of Biomarkers for Inflammation, Cardiac fibrosis and Cardiac Function: : Chronic inflammation will be quantified by serum levels of high sensitivity C-reactive protein (hsCRP) using a chemiluminescent immunometric assay (Immulite® 2500) at 0,20 and 40 weeks. Serum samples stored at -700 C and batch analysed for PICP (RIA- Framos Diagnostica), MMP1 and TIMP 1 (ELISA- Amersham) at University of Pamplona and Plasma NT proBNP will be measured by the Elecsys® proBNP immunoassay on the Roche Elecsys® 1010 immunoassay analyzer at baseline and at study completion.

(vii) Potential confounders: Haemoglobin will be maintained between 10-13gm%. Use of Erythropoietin stimulating agents, dose/frequency of iron replacement and any blood transfusion will be documented. Office blood pressure as per BHS guidelines will be documented at vitamin D administration stages. Changes in anti-hypertensive medications over the follow-up period will be noted. In addition the following parameters will be assessed at 0, 20 and 40 weeks:

• Plasma Aldosterone/renin activity ratio;
• Pulse wave velocity, Central Arterial Pressure and Augmentation Index (SphygmoCor Pulse Wave Velocity system);
• Serum calcium x phosphate product (<5 mmol/L);
• Urinary protein-creatinine ratio, sodium and fractional excretion of phosphate;
• Serum FGF-23.

(viii) Creation of a Biobank 10ml EDTA samples for cells and 3x3 ml tempus tubes for mRNA on PBMC/whole blood will be collected at week 0, 20 and 40 weeks for creation of a biobank at Kings college London (Guy's Campus) for future analysis of novel genes or biological pathways.

(ix) Recruitment and retention of trial participants

Planning and preparation of the trial would be conducted over the first 6 weeks. 120 CKD patients/week are seen at the three hospital sites. 30% of the patients are diabetic and 60% of the patients have a GFR < 45 mL/min/1.73m2. The investigators estimate vitamin D deficiency to be about 35% based on our pilot study. Recruitment will be over 50 weeks. The total duration of the trial will be 102 weeks. A timeline of the proposed study is described in figure 3.

(7) Power calculation and statistical analysis The interstudy variability of CMR for LV mass using a protocol similar to ours, has been reported between 2.8% to 4.8% with a SD of 8.4g(23). To demonstrate a 10g improvement in LV mass in the active treatment group, the investigators would need 19 patients [α= 0.05 and Power=95%]. Accounting for a dropout rate of 30%, the investigators will recruit 25 patients in each group. Comparisons for continuous variables will be performed using independent sample t-test between the active treatment and placebo group.

(8) Expected value of results The investigators are confident that the study is adequately powered to show a 10 gm change in LVM, using MR scan at Reza Rezavi's laboratory, if it exists; similarly a change in systolic function using echocardiogram at the expert laboratory at St Georges. The investigators will also be able to examine the role of fibrosis in the LVH and systolic dysfunction by measuring biomarkers at Javier Diez's laboratory. A large CKD patient population at St George's, Guy's and Kingston Hospitals will facilitate recruitment.

Figure 3: Timeline expressed in weeks showing planning, recruitment and anticipated randomization of 50 patients with 25 (OH) vitamin D levels < 37.5 nmol/L. Cardiac magnetic resonance, Echocardiography, markers of cardiac fibrosis will be measured at randomization and at study completion. CMR- Cardiac magnetic resonance, TDI- Tissue Doppler imaging, TTE- Transthoracic Echocardiography. Analysis and reporting of findings will take place between 96-102 weeks.

(9) Potential problems and alternative strategies:

• Impact of potential confounders has been minimised through strict entry criteria and follow-up protocol. A 30% attrition rate has been allowed for in our sample size calculation to accommodate changes in patients' clinical status that may account for observed changes in cardiac measurements, such as rapid decline in renal function, MI or death during trial follow-up, pre-emptive kidney transplantation, CaxP product > 5mmol/L, Hb < 10 or > 13gm% and persistent high office BP(> 150/90mmHg).
• Although patients in our pilot study were young, the investigators expect proportionate improvement in parameters in the trial participants as well.
• The investigators are conscious that improvement in LV function and morphology has not been previously shown with vitamin D therapy in human subjects. However, evidence of genomic effects of vitamin D on cardiac myocytes, extracellular matrix and myocardial contractility are well established in animal studies. Most importantly, this study explores a novel anti-fibrotic mechanism of vitamin D therapy in humans, which could provide the proof of principle for use of vitamin D in other disease states such as autoimmune disorders.

Inclusion Criteria:

• 50 adult (18-75 years old)
• patients with estimated GFR between 15-45 ml/min/1.73 m2 (CKD stage 3b and 4) on ACE inhibitors and/or ARB therapy will be recruited.

Exclusion Criteria:

The following patients will be excluded from the trial:

• patients with diabetes mellitus
• Hb < 10gm% or on EPO
• serum calcium > 2.55 mmol/L
• known malignancies
• patients on vitamin D supplementation
• heart failure (ejection fraction < 40 or pro NT BNP > 500 pg/ml
• uncontrolled hypertension (BP > 150/90 mmHg)
• patients with AV fistula for dialysis access
• significant valvular heart disease
• conditions that may influence collagen metabolism such as recent (< 6months) surgery or trauma
• fibrotic diseases or active inflammatory conditions
• history of myocardial infarction and/or Troponin T > 0.5