Effects of Myo-inositol Hexaphosphate (SNF472) on Bone Mineral Density in Patients Receiving Hemodialysis

An Analysis of the Randomized, Placebo-Controlled CaLIPSO Study

Abstract

Background and objectives 

In the CaLIPSO study, intravenous administration of SNF472 (300 or 600 mg) during hemodialysis significantly attenuated progression of coronary artery and aortic valve calcification. SNF472 selectively inhibits formation of hydroxyapatite, the final step in cardiovascular calcification. Because bone mineral is predominantly hydroxyapatite, we assessed changes in bone mineral density in CaLIPSO.

Design, setting, participants, & measurements 

Patients with coronary artery calcification at screening (Agatston score of 100–3500 U) were randomized 1:1:1 to receive placebo, 300 mg SNF472, or 600 mg SNF472 as an intravenous infusion during hemodialysis three times weekly for 52 weeks. Dual-energy x-ray absorptiometry (DXA) scans were obtained at baseline (screening) and end of treatment, and between-group changes from baseline were compared using analysis of covariance.

Results 

Among 274 randomized patients, 202 had evaluable DXA scans at baseline and postrandomization (the DXA-modified intention-to-treat population). Mean (95% confidence interval) changes in total-hip bone mineral density from baseline to week 52 were −1.5% (−2.7% to −0.3%), −1.5% (−2.7% to −0.4%), and −2.5% (−3.8% to −1.2%) in the placebo, 300 mg SNF472, and 600 mg SNF472 groups, respectively. Mean (95% confidence interval) changes in femoral-neck bone mineral density from baseline to week 52 were −0.3% (−1.6% to 1.0%), −1.0% (−2.3% to 0.2%), and −2.6% (−4.0% to −1.3%), respectively. Regression analyses showed no correlation between change in coronary artery calcium volume and change in bone mineral density at either location. Changes in serum alkaline phosphatase, calcium, magnesium, phosphate, and intact parathyroid hormone levels were similar across treatment groups. Clinical fracture events were reported for four of 90, three of 92, and six of 91 patients in the placebo, 300 mg SNF472, and 600 mg SNF472 groups, respectively.

Conclusions 

Bone mineral density decreased modestly in all groups over 1 year. In the 600 mg SNF472 group, the reduction appeared more pronounced. Reported fractures were infrequent in all groups.

Clinical Trial registry name and registration number: 

Effect of SNF472 on Progression of Cardiovascular Calcification in End-Stage-Renal-Disease (ESRD) Patients on Hemodialysis (HD), NCT02966028

Visual Abstract

Introduction

Patients with kidney failure have all-cause and cardiovascular mortality rates that are up to 30 times higher than the general age- and sex-matched population (^12–3). Cardiovascular causes are responsible for approximately half of deaths among patients receiving dialysis (⁴). More than 80% of patients receiving dialysis have evidence of cardiovascular calcification (⁵,⁶); these patients have two- to five-fold more cardiovascular calcification than age-matched individuals with established coronary artery disease (⁷). Intimal calcification (a marker of calcification of atherosclerotic plaque) is accelerated in patients receiving dialysis, as is medial calcification (Mönckeberg medial sclerosis), which exacerbates vascular stiffening and arteriosclerosis (⁸,⁹).

Several clinical studies have been conducted over the past two decades to identify treatments that slow the progression of cardiovascular calcification, ultimately aiming to reduce the frequency of cardiovascular events in patients with CKD. Most attempts have focused on control of serum phosphate or parathyroid hormone as targets of therapy (^1011–12). Some approaches have attenuated progression of cardiovascular calcification, but no approach to date has targeted the calcification process directly. SNF472, an intravenous formulation of the hexasodium salt of myoinositol hexaphosphate (phytate; IP6), selectively and directly inhibits the formation and growth of hydroxyapatite crystals, the final common step in the development of cardiovascular calcification (¹³,¹⁴).

We recently reported the effects of SNF472 on cardiovascular calcification in the multicenter, double-blind, placebo-controlled, randomized clinical study CaLIPSO (Cal for calcium and ipso meaning the item itself) (¹⁵), demonstrating that administration of SNF472 during each hemodialysis session for 1 year significantly attenuates progression of coronary artery and aortic valve calcification (¹⁶). Contemporary standard care was continued throughout the study in all treatment groups; therefore, the treatment effect of SNF472 was evaluated in addition to the effect of concomitant medications known to attenuate cardiovascular calcification, including noncalcium-containing phosphate binders and cinacalcet (¹²,¹⁷).

Patients with kidney failure have clinically meaningful abnormalities in bone related to changes in serum calcium, phosphate, parathyroid hormone, and circulating levels of vitamin D, induced by the reduction in glomerular filtration. These ionic, hormonal, and bone abnormalities are termed CKD–mineral and bone disorder (¹⁸). The abnormal bone in kidney failure is associated with higher incidence of fractures (¹⁹) and mortality (²⁰). As in individuals with normal kidney function, a reduction in bone mineral density (BMD) in patients with CKD is associated with a higher risk for subsequent fracture (²¹), but this association is not as clear in patients with kidney failure.

Bone mineral is predominantly hydroxyapatite, which is continually undergoing formation and resorption (²²,²³). Because SNF472 directly inhibits the formation of hydroxyapatite crystals and slows progression of cardiovascular calcification, we also assessed the effects of SNF472 on BMD. Preclinical studies showed no deleterious effects of SNF472 on bone mineralization in dogs and no impairment of rat osteoblasts (²⁴), but the effects of SNF472 on bone in humans have not been reported previously. We assessed BMD in CaLIPSO as a secondary safety end point, using dual-energy x-ray absorptiometry (DXA) scans of the total hip and femoral neck. In this report, we describe changes from baseline in BMD after 52 weeks of treatment with SNF472 versus placebo.

Materials and Methods

Overall Study Design

The study design was reported previously (¹⁵). Briefly, CaLIPSO was a multinational, randomized, double-blind, placebo-controlled, phase 2b clinical trial at 65 centers in three countries (United States, Spain, and United Kingdom) that investigated the effect of SNF472 on cardiovascular calcification in adult patients receiving maintenance hemodialysis. Patients with an Agatston coronary artery calcium score of 100–3500 U were randomized 1:1:1 to receive placebo, 300 mg SNF472, or 600 mg SNF472 as an intravenous infusion via the hemodialysis lines three times weekly for 52 weeks. To maintain blinding of study centers and patients, randomization was performed with a centralized, electronic randomization system. The study drug was packaged in identical vials containing either SNF472 or physiologic saline. Randomization was stratified by baseline coronary artery calcium Agatston score (100–399, 400–1000, or >1000 U). Enrollment began in December 2016 and continued through July 2018. Follow-up ended in August 2019.

At baseline and end of study (week 52 or early termination), multidetector computed tomography scans were conducted to analyze the primary efficacy end point of coronary artery calcium volume score, and the secondary efficacy end points of coronary artery calcium (Agatston), aortic valve calcium (volume and Agatston), and thoracic aorta calcium (volume and Agatston).

Safety measurements included adverse events at any time during the study, laboratory values at selected visits, and BMD measurements at baseline and end of study (see the next section for details). An external, independent data and safety monitoring board monitored patient safety and data integrity throughout the study. The data and safety monitoring board did not identify any safety signals (¹⁶).

The trial was conducted according to the principles of the Declaration of Helsinki, and ethical approval was obtained by the local institutional review boards at each site in accordance with local/national processes. Participants provided written informed consent before enrollment. The trial was registered on November 17, 2016.

Bone Mineral Density Measurements

We measured total-hip and femoral-neck BMD (mainly cortical bone) by DXA using either a Hologic or Lunar scanner for both baseline and end of study (week 52 or early termination); the same scanner was used at both assessments. We ensured consistency in imaging modality, anatomic positioning, coverage, and parameters across both imaging visits for each patient. For the Hologic scanner, we used the array mode for each image; the Prodigy Lunar scanner automatically chooses thin (<13 cm), standard (13–25 cm), or thick (>25 cm) cuts, depending on the body thickness and height/weight of the patient. We did not measure BMD of the spine (including mainly trabecular bone and a thin layer of cortex) because of the known influence of vascular (aortic) calcification on lumbar BMD measurements (²⁵,²⁶). A central imaging laboratory, which was blinded to treatment arm, confirmed the quality of all DXA images. An independent DXA expert reviewed and confirmed that the regions of interest were accurately identified and adjusted if required. Clinical fractures were reported by investigators. Fracture events were not adjudicated. No screening imaging for subclinical fractures was conducted.

Statistical Analyses

The DXA-modified intention-to-treat population included patients who were randomized, received at least one dose of study treatment (SNF472 or placebo), and had an evaluable DXA scan at baseline and postbaseline (week 52 or early termination). If a patient had an evaluable postbaseline scan at the early termination visit, we used that last observation carried forward for the DXA-modified intention-to-treat population. The DXA per-protocol population included randomized patients who had evaluable BMD scans at baseline and week 52, met all inclusion and exclusion criteria, and received at least 80% of study treatment. This approach was consistent with the definitions for the modified intention-to-treat and per-protocol populations for efficacy (¹⁶). The statistical analysis plan prespecified the primary analysis of BMD to be consistent with the efficacy analyses (i.e., comparing the combined SNF472 groups versus placebo). Analyses of individual dose groups versus placebo and of the per-protocol population were considered exploratory. The safety population for adverse event analyses included all patients who received at least one dose of study treatment. The planned sample size of 270 patients provided 80% power for the comparison of the combined SNF472 dosing groups and the placebo group for the primary study end point. The study was not powered for comparisons of DXA across study treatments.

To determine change in BMD, we used analysis of covariance with the change in log BMD (log [week 52]–log [baseline]) as the dependent variable and with a fixed-effect term for treatment group and log (BMD) as covariate, stratified by baseline coronary artery calcium Agatston score category (100–399, 400–1000, or >1000 U). We estimated geometric least squares means and 95% confidence intervals (95% CIs) and back-transformed them to yield the mean percent change from baseline to week 52. For each comparison between SNF472 and placebo, we calculated the geometric least squares mean and 95% CI for treatment differences. We also conducted a multivariable analysis with relevant covariates that were imbalanced across treatment groups at baseline. We created waterfall plots of change in BMD for total hip and femoral neck to graphically display individual changes in BMD. We examined scatterplots with regression lines to analyze the relation between change in BMD and change in coronary artery calcium volume. We summarized changes in laboratory assessments relevant to bone metabolism (serum albumin, alkaline phosphatase, calcium, magnesium, phosphate, and intact parathyroid hormone).

Results

A total of 273 patients were randomized to the three treatment groups and received at least one dose of study treatment. The DXA-modified intention-to-treat population included 202 patients, and the DXA per-protocol population included 142 patients (Figure 1).

Figure 1.: Patient disposition and assessment populations. *One patient in the placebo group with evaluable DXA of the femoral neck at baseline and postbaseline had evaluable DXA of the total hip only at postbaseline. DXA, dual-energy x-ray absorptiometry; mITT, modified intention to treat.

Baseline characteristics of patients in the DXA-modified intention-to-treat population are shown in Table 1. Some imbalance was noted in baseline variables that may influence bone, including dialysis vintage at baseline, concomitant use of calcimimetics, and concomitant use of activated vitamin D. Baseline values for BMD were similar across the treatment groups, both in the total hip and in the femoral neck. Patients included in the DXA-modified intention-to-treat population were more likely to be White, and patients not included in the DXA-modified intention-to-treat population were more likely to be Black; other baseline characteristics were similar between the two populations overall (Supplemental Table 1).

Table 1. - Demographic and baseline characteristics (DXA-modified intention-to-treat population)

Characteristics	Placebo (n=67)	300 mg SNF472 (n=75)	600 mg SNF472 (n=60)
Age, yr
Mean±SD	64±8	63±10	64±10
Median (range)	65 (35–79)	63 (41–83)	63 (36–79)
Age >55 yr, n (%)	56 (84)	58 (77)	51 (85)
Female, n (%)	22 (33)	30 (40)	23 (38)
Hispanic or Latino, n (%)	25 (37)	25 (33)	26 (43)
Race, n (%) a
White	49 (73)	51 (68)	47 (78)
Black	13 (19)	19 (25)	7 (12)
Asian	3 (4)	2 (3)	3 (5)
Other	0 (0)	1 (1)	1 (2)
Not reported	2 (3)	3 (4)	2 (3)
Body mass index
Mean±SD (kg/m2)	29.8±6.2	28.4±6.3	28.5±5.8
>19 kg/m2, n (%)	66 (99)	71 (95)	58 (97)
Dialysis vintage, mo
Mean±SD	44±38	60±49	56±55
Median (range)	30 (6–150)	50 (7–188)	42 (7–339)
Relevant medical history, n (%)
Osteoporosis	3 (4)	4 (5)	3 (5)
Fractures	5 (7)	2 (3)	3 (5)
Parathyroidectomy	1 (1)	1 (1)	1 (2)
Hyperparathyroidism	42 (63)	57 (76)	37 (62)
Concomitant medication use, n (%)
Activated vitamin D	41 (61)	53 (71)	31 (52)
Bisphosphonates	0 (0)	1 (1)	2 (3)
Calcimimetics	21 (31)	33 (44)	16 (27)
Calcium supplementation	26 (39)	29 (39)	26 (43)
Noncalcium-based phosphate binders	45 (67)	49 (65)	43 (72)
Warfarin	4 (6)	10 (13)	7 (12)
BMD at baseline, mean±SD
Total hip (g/cm2)	0.839±0.168 b	0.841±0.182	0.802±0.155
Femoral neck (g/cm2)	0.745±0.172	0.760±0.153	0.723±0.147

DXA, dual-energy x-ray absorptiometry; BMD, bone mineral density.

^aA patient could select more than one option for race.

^bIn the placebo group, total-hip BMD was evaluable at baseline for 66 patients.

Table 2 shows changes in geometric least squares means from baseline to week 52 for BMD in the total hip and femoral neck. In the 600-mg group, the observed reductions in BMD appeared more pronounced. Results were similar when analyses were covariate adjusted. Differences in BMD were attenuated in the per-protocol population.

Table 2. - Bone mineral density: change from baseline to week 52

Population/Statistic	Placebo	Combined Dosing Groups	300 mg SNF472	600 mg SNF472
Total hip
DXA mITT population, primary analysis
No. of patients	66	135 a	75 a	60
Change, % (95% CI)	−1.5 (−2.7 to −0.3)	−2.0 (−2.9 to −1.2)	−1.5 (−2.7 to −0.4)	−2.5 (−3.8 to −1.2)
DXA mITT population, adjusted for covariates b
No. of patients	66	135 a	75 a	60
Change, % (95% CI)	−1.4 (−2.7 to −0.1)	−2.0 (−3.0 to −1.1)	−1.6 (−2.8 to −0.4)	−2.4 (−3.8 to −1.0)
DXA per-protocol population
No. of patients	43	99	55	44
Change, % (95% CI)	−2.0 (−3.5 to −0.6)	−1.6 (−2.6 to −0.7)	−0.8 (−2.1 to 0.4)	−2.4 (−3.8 to −1.0)
Femoral neck
DXA mITT population, primary analysis
No. of patients	67	135 a	75 a	60
Change, % (95% CI)	−0.3 (−1.6 to 1.0)	−1.8 (−2.8 to −0.9)	−1.0 (−2.3 to 0.2)	−2.6 (−4.0 to −1.3)
DXA mITT population, adjusted for covariates b
No. of patients	67	135 a	75 a	60
Change, % (95% CI)	−0.2 (−1.6 to 1.2)	−1.8 (−2.8 to −0.7)	−1.0 (−2.3 to 0.3)	−2.5 (−4.0 to −1.0)
DXA per-protocol population
No. of patients	43	99	55	44
Change, % (95% CI)	−0.9 (−2.6 to 0.8)	−1.1 (−2.2 to 0.0)	−0.5 (−2.0 to 1.0)	−1.7 (−3.3 to −0.1)

DXA, dual-energy x-ray absorptiometry; mITT, modified intention-to-treat; 95% CI, 95% confidence interval.

^aThe change from baseline to week 52 could not be analyzed for one subject in the 300 mg SNF472 group.

^bCovariates added: dialysis vintage at baseline, concomitant use of calcimimetics, and concomitant use of activated vitamin D.

Waterfall plots of the change in total-hip and femoral-neck BMD for individual patients (Figure 2) showed a similar pattern of the change from baseline in BMD in all groups.

Figure 2.: Waterfall plots of change from baseline in total-hip and femoral-neck bone mineral density (DXA - modified intention-to-treat population). (A-D) total hip; (E-H) femoral neck. There was a similar pattern of the change from baseline in BMD in all groups. BMD, bone mineral density; ET, early termination.

Regression analyses showed no correlation (inverse or direct) between the change from baseline in coronary artery calcium volume and the change from baseline in BMD in any group (Figure 3).

Figure 3.: Regression analysis of change in bone mineral density versus change in coronary artery calcium (CAC) volume (DXA- modified intention-to-treat population). There was no correlation (inverse or direct) between the change from baseline in coronary artery calcium volume and the change from baseline in BMD in any group. Adj, adjusted.

Changes from baseline to week 52 for laboratory assessments relevant to bone metabolism are summarized in Table 3. Mean and median changes from baseline to the end of treatment for these assessments were similar across the treatment groups and were small in magnitude. However, the variability was relatively large, particularly for levels of serum alkaline phosphatase and parathyroid hormone.

Table 3. - Laboratory assessments relevant to bone metabolism: change from baseline to week 52 (DXA-modified intention-to-treat population)

Laboratory Assessment	Placebo (n=67)	Combined Dosing Groups (n=135)	300 mg SNF472 (n=75)	600 mg SNF472 (n=60)
Albumin (g/dl)
N	55	119	68	51
Mean±SD	0.04±0 29	−0.01±0.30	0.02±0.31	−0.04±0.28
Median	0.0	0.0	0.0	0.0
Q1, Q3	−0.1, 0.2	−0.2, 0.2	−0.2, 0.3	−0.2, 0.2
Range	−0.5 to 0.7	−0.7 to 0.8	−0.6 to 0.8	−0.7 to 0.6
Alkaline phosphatase (U/L)
N	55	117	67	50
Mean±SD	−1.9±44.3	−1.5±51.6	−0.9±45.2	−2.4±59.5
Median	1.0	−1.0	0.0	−3.5
Q1, Q3	−23.0, 20.0	−26.0, 18.0	−28.0, 21.0	−23.0, 14.0
Range	−128 to 175	−237 to 224	−106 to 145	−237 to 224
Calcium (mg/dl)
N	55	118	68	50
Mean±SD	0.20±1.0	0.08±0.72	0.04±0.72	0.16±0.68
Median	0.24	0.20	0.20	0.24
Q1, Q3	−0.28, 0.64	−0.40, 0.52	−0.38, 0.52	−0.40, 0.60
Range	−3.7 to 3.4	−1.5 to 2.1	−1.5 to 2.1	−1.5 to 1.7
Magnesium (mg/dl)
N	55	118	68	50
Mean±SD	−0.05±0.32	−0.07±0.36	−0.10±0.41	−0.02±0.34
Median	0.02	−0.10	−0.10	−0.07
Q1, Q3	−0.10, 0.17	−0.24, 0.10	−0.28, 0.09	−0.22, 0.19
Range	−1.3 to 0.61	−1.0 to 2.1	−0.97 to 2.1	−1.0 to 1.2
Phosphate (mg/dl)
N	21	36	18	18
Mean±SD	−0.50±1.83	0.00±1.36	−0.40±1.39	0.40±1.27
Median	−0.34	0.03	−0.37	0.31
Q1, Q3	−1.83, 0.46	−0.62, 0.79	−1.11, 0.62	−0.19, 1.21
Range	−3.9 to 3.8	−3.6 to 2.4	−3.6 to 2.4	−3.5 to 2.0
Intact PTH (pg/ml)
N	53	114	66	48
Mean±SD	−62±324	−43±421	−31±424	−58±422
Median	−24	−30	−31	−28
Q1, Q3	−259, 167	−167, 73	−177, 88	−136, 48
Range	−1003 to 698	−2315 to 1344	−1604 to 1344	−2315 to 857

DXA, dual-energy x-ray absorptiometry; Q1, first quartile (25th percentile); Q3, third quartile (75th percentile); PTH, parathyroid hormone.

Adverse events of fracture among all 273 patients are summarized in Table 4. Clinical fractures were reported for four of 90, three of 92, and six of 91 patients in the placebo, 300 mg SNF472, and 600 mg SNF472 groups, respectively. Fractures were not adjudicated and there was no pattern with respect to site of fracture.

Table 4. - Patient incidence of fracture (safety population)

Incidence of Fracture	n
	Placebo (n=90)	Combined Dosing Groups (n=183)	300 mg SNF472 (n=92)	600 mg SNF472 (n=91)
Any fracture	4	9	3	6
Hip fracture	2	2	0	2
Spinal-compression fracture	1	1	0	1
Ankle fracture	0	1	1	0
Femoral-neck fracture	0	1	1	0
Foot fracture	0	1	0	1
Fractured sacrum	0	1	0	1
Humerus fracture	1	0	0	0
Lower-limb fracture	0	1	1	0
Upper-limb fracture	0	1	0	1

Discussion

In this multicenter, double-blind, placebo-controlled, randomized clinical study (CaLIPSO), we previously demonstrated that the administration of SNF472 during each hemodialysis session significantly attenuated the progression of coronary artery and aortic valve calcification. Herein, we describe changes in total-hip and femoral-neck BMD in placebo, pooled, and individual SNF472 dose groups in the modified intention-to-treat and per-protocol populations. We show similar changes in BMD in the prespecified comparison of combined groups with placebo, and a reduction in BMD in the 600 mg SNF472 group; the observed decline was less for the femoral neck in the per-protocol population.

Hydroxyapatite is a principal component of the mineral phase in both atherosclerotic plaque and bone. Calcific plaque formation and bone mineralization are both cell-mediated, organized, and highly regulated processes (²⁷). In both atherosclerotic calcifications and bone, there is initially deposition of an amorphous calcium phosphate that matures and transforms into crystalline hydroxyapatite (²⁸,²⁹). IP6 is an endogenous inhibitor of calcification that binds to the surface of hydroxyapatite crystals and prevents subsequent calcium and phosphate precipitation, limiting subsequent crystal growth (¹⁴,³⁰,³¹). Although IP6 is present in fibrous plants (such as vegetables, nuts, and grains), little of it is absorbed in the mammalian gastrointestinal tract. SNF472 is an intravenously administered formulation of IP6 that has been shown to inhibit cardiovascular calcification in animal models, and in patients receiving maintenance hemodialysis (¹⁶,³²,³³).

Patients with CKD and those on dialysis have markedly higher fracture rates compared with age- and sex-matched persons, and patients with fracture experience a higher risk of death (¹⁹,²⁰). Fracture rates are progressively higher at lower levels of kidney function, with a risk of skeletal fracture that is up to five times higher with a GFR of <15 versus >60 ml/min per 1.73 m² (²⁰). Among participants in the Dialysis Outcomes and Practice Patterns Study (DOPPS), 3% experienced a fracture, although the incidence varied across countries from 12 events per 1000 patient-years in Japan to 45 events per 1000 patient-years in Belgium (³⁴). The incidence rate in all countries exceeded those reported for the general population. Patients with fracture had a 3.7-fold higher rate of death compared with the overall DOPPS population (³⁴). In patients with CKD, at least four prospective cohort studies demonstrated that BMD is associated with fracture risk (^353637–38). In contrast, data addressing the association of BMD with fracture among patients receiving hemodialysis are limited and inconclusive.

After achieving peak bone mass, the continuous formation and resorption of bone helps maintain bone quality, and the maintenance of peak bone mass helps to prevent fracture (²³). There appears to be an association between BMD and vascular calcification in patients on dialysis (³⁹). Braun et al. (⁴⁰) and Raggi et al. (⁴¹) both demonstrated an inverse correlation between BMD and the extent of vascular calcification, and both lower BMD and more extensive vascular calcification are correlated with mortality. Patients with severe vascular calcification appear to have lower rates of bone turnover, with less osteoclastic resorption and less osteoblastic formation (³⁹).

DXA-derived BMD assesses bone mass; it does not assess bone quality. The strength of bone, indicating its resistance to fracture, is a function not only of bone density but also of bone quality, which is itself a function of bone composition and structure. Especially in patients on dialysis, bone quality may be adversely affected by marked changes in parathyroid hormone, vitamin D, phosphate, and calcium (¹⁸). In this study, changes in calcium, magnesium, phosphate, albumin, intact parathyroid hormone, and alkaline phosphatase were similar between the placebo and SNF472 groups. More detailed measures of bone quality were not available in the CaLIPSO trial. The waterfall plots indicate that most patients in both groups experienced little to no change in BMD. As would be expected over time, a larger proportion of patients (in placebo and SNF472 groups) experienced a reduction, rather than an increase, in BMD. Because bone turnover is a relatively slow process, changes in BMD, and especially rates of fracture, are generally assessed over multiyear studies. Our study duration of 1 year limits the ability to determine the clinical significance of the changes observed. It is worth emphasizing that DXA cannot distinguish between cortical and trabecular BMD; fracture risk may rise differentially with reduction in cortical and trabecular BMD at different anatomic sites.

We previously reported that SNF472 significantly reduces progression of coronary artery and aortic valve calcification in patients receiving maintenance hemodialysis. As part of our safety evaluation, we examined changes in total-hip and femoral-neck BMD. The prespecified comparison showed similar changes from baseline between combined treatment groups and placebo. In the 600 mg SNF472 group, the reductions appeared larger. The reported fractures were infrequent. The clinical significance of this finding is not certain. There was no correlation between the change in coronary artery calcium and BMD. Future studies with SNF472 will continue to monitor adverse events, including those related to bone, to evaluate potential risks and benefits of this agent.

Disclosures

A. Bellasi reports receiving lecture fees Abbvie, Amgen, Sanofi-Genzyme, and Vifor-Fresenius-Renal Pharma; receiving honoraria from Amgen, Genzyme, Keryx, Sanifit, Sanofi, and Shire; receiving unrestricted research grants from Amgen and the Italian Society of Nephrology; serving as an associate editor for Atherosclerosis, Blood Purification, BMC Nephrology, Cardiorenal Medicine, and Journal of Nephrology; serving as a member of the European Renal Association–European Dialysis and Transplant Association and the Italian Society of Nephrology; having consultancy agreements with Sanifit; and being employed by Papa Giovanni XXIII Hospital, Bergamo, Italy. J. Bover reports serving as a consultant to, and receiving lecture fees from, Abbvie, Amgen, Rubio, Sanifit, Sanofi, and Vifor-Fresenius-Renal Pharma; receiving honoraria from Abbvie, Amgen, Rubio, Sanifit, Sanofi, and Vifor-Fresenius-Renal Pharma; serving on the medical advisory boards for Abbvie and Vifor-Fresenius-Renal Pharma; being employed by Puigvert Foundation; receiving research funding from Rubio; and receiving lecture fees from Shire. D. Bushinsky reports having consultancy agreements with Amgen, Ardelyx, Relypsa/Vifor/Fresenius, Sanifit, Sanofi, and Tricidia; serving as a scientific advisor for, or member of, Amgen, Ardelyx, Sanifit, Sanofi, Tricidia, and Vifor/Relypsa; receiving honoraria from Amgen, Ardelyx, Sanifit, Sanofi, Tricidia, and Vifor/Relypsa; having stocks or options in Amgen and Tricida; receiving research funding from National Institutes of Health and Renal Research Institute; serving on the speakers bureau for Sanofi; and being employed by University of Rochester Medical Center. G. Chertow reports having consultancy agreements with Akebia, Amgen, Ardelyx, AstraZeneca, Baxter, Cricket, DiaMedica, Gilead, Miromatrix, Reata, Sanifit, and Vertex; receiving research funding from Amgen, Janssen, and National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), and National Institute of Allergy and Infectious Diseases; being on the data safety monitoring board for Angion, Bayer, NIDDK, and ReCor; having stocks or options in Ardelyx, CloudCath, Cricket, Durect, DxNow, Eliaz Therapeutics, Outset, Physiowave, and PuraCath; serving as coeditor of Brenner & Rector’s The Kidney (Elsevier); serving on the board of directors for Satellite Healthcare; and being employed by Stanford University School of Medicine. R. Garg reports being employed by PharmaDRS Consulting, LLC; and having stock options in, and being a former employee of, Sanifit Therapeutics. A. Gold reports being employed by, and having stock options in, Sanifit Therapeutics. M. Ketteler reports having consultancy agreements with Amgen, AstraZeneca, Baxter, Boehringer Ingelheim, Medice, Sanifit, Sanofi, Shire-Takeda, and Vifor Pharma; serving on a speakers bureau for Amgen, AstraZeneca, Bristol-Myers Squibb, Boehringer Ingelheim, Fresenius Medical Care, Kyowa Kirin, Medice, Ono Pharmaceuticals, Pfizer, Shire-Takeda, and Vifor Pharma; serving as a scientific advisor for, or member of, the Kuratorium für Heimdialyse e.V. (KfH) Board of Directors, Kidney Disease Improving Global Outcomes Executive Committee, and Nephrology Dialysis Transplantation; and being employed by Robert Bosch Hospital, Stuttgart, Germany. J. Perelló reports having other interests/relationships with Biotech Cluster of the Balearic Islands and patents related to SNF472; and having employment with, consultancy agreements with, stock options in, receiving research funding from, receiving honoraria from, and serving as a scientific advisor or member of Sanifit Therapeutics. P. Raggi reports having consultancy agreements with Sanifit and being employed by the University of Alberta. M. Rodriguez reports receiving research funding from Amgen; serving as a scientific advisor or member of Kiowa and Sanifit; receiving honoraria from Amgen, Kyowa, Rubio, and Vifor; receiving lecture fees from Amgen, Kyowa Kirin, Sanofi, and Vifor; having consultancy agreements with Kyowa and Sanifit; and being employed by the University of Córdoba, Spain. S. Sinha reports receiving research funding from Amgen, Ethicon, and AstraZeneca; having consultancy agreements with Sanifit; receiving honoraria from Napp Pharmaceuticals, Novartis, Travere, and Vifor Pharma; and being employed by the Salford Royal National Health Service Foundation Trust.

Funding

This study was funded by Sanifit Therapeutics.