A phase 1b randomized, placebo‐controlled clinical trial with SNF472 in haemodialysis patients

What is already known about this subject

• Survival in haemodialysis (HD) patients is correlated with cardiovascular calcification, but no drug for the treatment of cardiovascular calcification has been approved to date. SNF472 is a new calcification inhibitor with reported efficacy in preclinical models. A recent study showed acceptable safety and tolerability in healthy volunteers and after single administration in HD patients.

What this study adds

• This is a double‐blind, randomised, placebo‐controlled study to assess the safety, pharmacokinetics and pharmacodynamics of intravenous SNF472 with repeated administrations to HD patients for up to 28 days. Our results support the safety and tolerability of SNF472 at repeated dosing for 4 weeks and show potential pharmacokinetic‐dependent SNF472 efficacy against calcification, thus providing a strong rationale for further clinical development.

1. INTRODUCTION

Cardiovascular calcification (CVC) is a significant risk factor for morbidity and mortality in patients with end‐stage renal disease (ESRD) on maintenance dialysis.1 Approximately 50% of patients with chronic kidney disease die from cardiovascular events.2 A correlation between the degree and progression of CVC and cardiovascular events (including mortality) has been demonstrated.3, 4, 5, 6 Key risk factors for the development of CVC in patients on haemodialysis (HD) include hyperphosphataemia, hypercalcaemia, hyperparathyroidism and abnormal vitamin D metabolism. An imbalance between inhibitors and enhancers of CVC are central to the basic pathophysiology of this condition.7, 8 Two of the major families of drugs used in HD patients each address 1 of the many factors that can lead to vascular calcification. Phosphate binders (such as sevelamer) address the hyperphosphataemia,9 whereas calcimimetics (i.e. cinacalcet) address the hyperparathyroidism10 and consequently the hypercalcaemia. However, none of these treatments target the common final pathway in the development of CVC.

SNF472 is a selective calcification inhibitor being investigated as a new approach for the treatment of CVC in HD patients with ESRD and for the treatment of calciphylaxis. It is an intravenous (IV) formulation of myo‐inositol hexaphosphate (IP6). In vitro11 and in vivo12, 13 results show that SNF472 or IP6 prevent the formation of hydroxyapatite (HAP) crystals. SNF472 directly inhibits calcium crystal formation by binding to the surface of the crystal, and IV SNF472 administration to HD patients could mitigate the impact of the imbalance between CVC inhibitors and enhancers present in these patients. By inhibiting the final common pathway of CVC, the efficacy of SNF472 is likely to be independent of the aetiology of CVC and would be observed irrespective of plasma calcium or phosphate levels.14, 15

The safety, tolerability and pharmacokinetics (PK) of IV SNF472 was evaluated in healthy volunteers (HV) following administration of single IV doses ranging from 0.5 to 12.5 mg/kg and in HD patients after treatment with 1 simple dose.16 In that Phase I first‐time‐in‐human clinical study, the PK profile showed a C_max proportionality and a relatively short half‐life. SNF472 was well tolerated with the exception of mild‐to‐moderate local infusion site irritation in HV.

The study reported in this paper was a Phase 1b, double‐blind, placebo‐controlled, randomized, multiple ascending dose clinical trial to evaluate the safety, tolerability and PK/pharmacodynamics (PD) of IV SNF472 in HD patients.

2. METHODS

This study was conducted at a single site in accordance with the International Conference for Harmonization E6 Guidelines for Good Clinical Practice and the Declaration of Helsinki. This study is registered at Eudra CT (Identifier 2014–005076‐29) and was approved by an Independent Ethics Committee. Written informed consent was obtained from all patients prior to the start of the study.

2.1. Patients

Patients were male or nonpregnant female stable HD patients, aged 18–80 years, inclusive. At screening, patients should have been undergoing standard HD 3 times a week for at least 6 months. Causes of exclusion included: alkaline phosphatase ≥240 IU/L; hospitalization within the previous 3 months for myocardial ischaemia, unstable angina, heart failure, peripheral vascular disease, stroke or seizures; a schedule date for a renal transplant; positive testing for hepatitis B or human immunodeficiency virus; history of drug abuse in the past 2 years or participation in another trial within 1 month prior to the first dose of study drug.

2.2. Study design

The study had 2 subject cohorts with 8 patients per cohort (Figure 1 ). Cohort 1 received multiple ascending doses of study drug for 1 week, and Cohort 2 received a fixed repeated dose for 4 weeks.

Figure 1.

Flowchart of distribution of patients in Cohort 1 A, and Cohort 2 B

Cohort 1 consisted of 10 patients (8 were planned, but 2 additional patients were recruited to replace 2 patients that discontinued the study) and involved placebo and 5 different doses of study drug (Table S1 and Figure 1). Subjects participated in 5 treatment periods (TP) of 1 week each. For the 3 highest dose groups (TP1, TP2 and TP3), subjects were randomized in each TP in a 6:2 ratio to receive SNF472 (6 patients) or placebo (2 patients), ensuring that subjects who had received placebo in a specific TP received active treatment in the other 2 TPs. Therefore, a total of 6 different placebos were studied in Cohort 1. For the 2 lowest dose groups, the 8 patients from Cohort 1 were divided into 2 additional TPs (TP4 and TP5) of 4 patients each, after a washout of at least 1 month, with no additional placebos. The Dose Escalation Safety Committee (DESC) reviewed PK, tolerability, and safety data after each week of treatment in Cohort 1 (TP1 and TP2) to decide whether to proceed with dose escalation.

In cohort 2, subjects participated in a multiple dose study. They were randomized in a 6:2 ratio to receive SNF472 (6 patients) or placebo (2 patients). Patients received a single IV dose of either SNF472 or placebo on days 1, 3, 5, 8, 10, 12, 15, 17, 19, 22, 24 and 26 during the patient's dialysis session.

2.3. Treatments

The doses for Cohort 1 were 1, 3, 5, 12.5 and 20 mg/kg. For Cohort 2, the dose was 10 mg/kg. Study drug administration consisted of an IV dose of either SNF472 or placebo infused over a 4‐hour period in the morning during the subject's dialysis session. Subjects were administered study drug 3 times per week. The appropriate volume of the study drug was administered using a 100 mL bag of saline connected to an infusion pump, which was connected directly to the dialysis machine (Cordiax 5008; Fresenius Medical Care, USA) via an IV giving set (Sendal, Spain) and an accessory heparin line (Gambro Dasco, Italy; Figure S2).

2.4. Pharmacokinetic assessments

Serial blood samples for PK were obtained at specific time points (predose and up to 8 hours postdose) during the TP. Blood samples were collected into K₃EDTA anticoagulant tubes and were centrifuged at 1900 g for 10 minutes at 4°C. Plasma was stored at −80°C until IP6 quantification by UPLC‐MS using the validated method described by Tur et al.17

The PK parameters calculated were: area under the curve from 0 to the time of the last quantifiable concentration (AUC_0➔t); AUC from 0 to infinity (AUC_inf); percentage of AUC extrapolated from AUC_0➔t (%AUC_ex); total body clearance (CL); maximum SNF472 concentration measured in plasma (C_max); terminal elimination half‐life (t_1/2); time at which the C_max is measured (T_max); volume of distribution at terminal phase (V_z); and apparent first order terminal elimination rate constant (λ_z).

2.5. Pharmacodynamic assessments

The PD of SNF472 administered by IV infusion was evaluated by measuring HAP crystallization.11 For this purpose, 2 samples from each subject were used for PD analysis: 1 sample obtained at baseline before SNF472 infusion and 1 sample at T_max, obtained at the end of the 4 hours of infusion. Plasma samples were mixed in 0.15 M NaCl (pH 7.4), with 1.5 mM phosphate and 12.5 mM calcium in 96‐well plates, and the crystallization of HAP was measured spectrophotometrically at 550 nm for 30 minutes.

2.6. Safety assessments

Safety parameters included: treatment‐emergent adverse events (TEAEs)_; clinical laboratory and haematology_; clinical chemistry (including plasma electrolytes)_; vital signs (pulse, respiratory rate, supine blood pressure and axillary body temperature); 12‐lead electrocardiogram parameters (including cardiac intervals, PR, QRS, QT and QTc); and physical examination. Ionized calcium was measured at T_max in all patients to monitor patients for hypocalcaemia.

2.7. Statistical analyses

Data were collected in PostgreSQL by OpenClinica™. All statistical analyses and data manipulation were performed using SAS version 9.3 (SAS Institute Inc., Cary, NC, USA).

The PK analysis was performed using the PK population defined as all patients who received at least 1 dose of SNF472 and for whom either of the primary PK parameters (C_max or AUC_0‐t) could be calculated for at least 1 TP and for whom no major protocol deviation occurred. PK parameters of SNF472 were calculated by noncompartmental analysis methods from the concentration–time data. PK parameters were summarized descriptively using arithmetic mean, standard deviation, coefficient of variation (CV%), median, minimum, maximum, geometric mean, geometric CV% and number of observations.

To assess PD, inhibition of calcification induction was calculated as the percentage change in the slope for the linear interval between predose and postdose samples, using the following formula:

$$\%\mathit{\text{Inhibition}}=\left[\left(\mathit{\text{slope postdose}}–\mathit{\text{slope predose}}\right)/\mathit{\text{slope predose}}\right]*100$$

For Cohort 1, the statistical significance of the data was initially assessed by a 2‐way analysis of variance (ANOVA) considering the factors time (days) and dose (0, 1, 3, 5, 12.5 and 20 mg/kg). A significant effect of dose but not of time was obtained. Therefore, a 1‐way ANOVA was performed combining days 1 and 5 and considering only SNF472 dose as an independent variable. A posthoc analysis was applied to determine the differences between the groups involved. For Cohort 2, the statistical significance of the time‐course data was assessed by a 2‐way ANOVA considering the factors time (days) and treatment (placebo vs SNF472 10 mg/kg). A significant effect of treatment but not of time was obtained. Therefore, a Student t test was performed combining days 1, 8, 15 and 26, and considering only SNF472 treatment as a factor.

3. RESULTS

3.1. Patients and treatment

In Cohort 1, 6 patients completed all periods, 2 were prematurely withdrawn after completing TP1, leading to 2 replacement patients in TP2. In Cohort 2, 8 patients were included and all patients completed the study.

Demographic characteristics were similar for Cohort 1 and Cohort 2, with mean ages of 61.8 and 62.1 years, respectively, and mean BMI values of 28.6 and 27.5 kg/m², respectively (Table 1 ). The majority of patients were Caucasian–non‐Hispanic (70.0% in Cohort 1 and 62.5% Cohort 2) and male (60% in Cohort 1 and 87.5% in Cohort 2). A list of concomitant medication related to chronic kidney disease‐mineral bone disorder management is shown in Table S3.

Table 1.

Demographic characteristics and alcohol consumption habits

Variable	Statistic	Cohort 1 (n = 10)	Cohort 2 (n = 8)
Sex
	n	10	8
	Female	4 (40.0%)	1 (12.5%)
	Male	6 (60.0%)	7 (87.5%)
Age (y)	n	10	8
	Mean	61.8	62.1
	Standard deviation	15.1	10.1
	Median	65.5	66.0
Height (cm)	n	10	8
	Mean	161.8	169.6
	Standard deviation	5.9	10.3
	Median	161.0	172.0
Weight (kg)	n	10	8
	Mean	75.0	79.0
	Standard deviation	23.3	13.8
	Median	68.2	77.4
Body mass index (kg/m2)	n	10	8
	Mean	28.6	27.5
	Standard deviation	8.2	4.4
	Median	27.1	28.7
Race	n	10	8
	Caucasian– Hispanic	1 (10.0%)	3 (37.5%)
	Caucasian–non‐Hispanic	7 (70.0%)	5 (62.5%)
	Other	2 (20.0%)	0
Alcohol consumption	n	10	8
	Yes	1 (10.0%)	0
	No	9 (90.0%)	8 (100.0%)

3.2. PK

In cohort 1, C_max was observed at the end of the 4‐hour infusion (T_max; Figure 2A and Table 2) and concentrations declined rapidly falling below the lower limit of quantitation (LLOQ; 0.5 μg/mL) by 6 hours (5 mg/kg) and 8 hours (12.5, 20 mg/kg) postdose. At a dose of 1 mg/kg, only 1 subject had measurable plasma concentration.

Figure 2.

Pharmacokinetic profile of intravenous SNF472 in haemodialysis patients in a multiple ascending dose. A, Plasma SNF472 levels at Day 1. B, SNF472 plasma C_max at Day 1 and Day 5. Results represent mean ± standard deviation

Table 2.

Summary of pharmacokinetic parameters

Cohort 1
Parameter	SNF472 1 mg/kg Mean (CV%)	SNF472 3 mg/kg Mean (CV%)	SNF472 5 mg/kg Mean (CV%)	SNF472 12.5 mg/kg Mean (CV%)	SNF472 20 mg/kg Mean (CV%)
Day 1
Cmax (μg/mL)	‐	7.6 (66)	16.1 (33)	46.0 (34)	66.9 (43)
Tmax (min)	‐	236 (0)	217 (22)	200 (31)	244 (8)
AUC0‐t (μg h/mL)	‐	20.7 (72)	48.7 (25)	146.5 (37)	282.5 (49)
AUC0‐∞ (μg h/mL)	‐	1.4 (70)	51.8 (26)	158.0 (39)	350.2 (48)
t1/2 (min)	‐	34 (11)	32 (26)	52 (39)	65 (61)
Vz (mL)	‐	6536 (22)	6165 (46)	7973 (62)	5989 (29)
Vss (mL)	‐	11746 (34)	9395 (33)	8766 (65)	7295 (13)
CL (mL/h)	‐	16957 (86)	8759 (35)	6309 (32)	4662 (48)
Day 5
Cmax (μg/mL)	0.7a	8.0 (50)	19.3 (33)	46.8 (46)	56.0 (38)
Tmax (min)	232a	235 (1)	239 (1)	221 (23)	214 (22)
AUC0‐t (μg h/mL)	2.1a	24.4 (61)	57.0 (31)	185.6 (55)	206.2 (40)
AUC0‐∞ (μg h/mL)	2.3a	25.3 (61)	59.9 (36)	162.9 (46)	199.2 (46)
t1/2 (min)	NC	35 (34)	32 (53)	51 (32)	55 (53)
Vz (mL)	NC	6689 (33)	5772 (53)	7235 (33)	9691 (80)
Vss (mL)	NC	11356 (13)	9324 (32)	8927 (19)	9165 (39)
CL (mL/h)	33783a	10603 (48)	7603 (28)	6196 (29)	7476 (42)

Cohort 2
	SNF472 10 mg/kg
Parameter	Day 1	Day 26
Cmax (μg/mL)	38.7 (54)	41.5 (29)
Tmax (min)	238 (2)	237 (1)
AUC0‐t (μg h/mL)	115.2 (54)	104.2 (25)
AUC0‐∞ (μg h/mL)	120.4 (53)	106.1 (26)
t1/2 (min)	39 (21)	25 (15)
Vz (mL)	5695 (24)	4647 (25)
Vss (mL)	9415 (30)	8715 (28)
CL (mL/h)	10717 (100)	7945 (30)

AUC, Area under the curve; AUC_0‐t, Area Under the Curve from 0 to the time of the last quantifiable concentration; AUC_0‐∞, Area Under the Curve from 0 to infinity; CL, Clearance; C_max, Maximum SNF472 concentration measured in plasma; NC = not counted; t_1/2, Terminal elimination half‐life; T_max, Time at which the C_max is measured; Vz, Volume of distribution at terminal phase; Vss, Volume of distribution at steady state.

Note: T_max values correspond to the arithmetic mean of nominal times at which C_max was reached (just after infusion period).

‐: all patients (n = 6) presented plasma levels below of limit of detection (0.5 μg/ml).

^a n = 1.

The mean C_max, AUC_0‐t, and AUC_0‐∞ values of SNF472 increased with increasing doses of SNF472 on day 1 and day 5 of administration in a slightly more than dose‐proportional manner (Figure 2B). No significant differences were observed between the systemic exposure (C_max and AUC) achieved on day 1, day 3 (C_max), and day 5 of administration.

There was no apparent accumulation with multiple doses. Despite the short half‐life observed (32–65 min), steady state was not achieved during the 4‐hour infusion period. Based on a noncompartmental PK model, accumulation and volume of distribution calculations based on the distribution t_1/2 are shown in Table 2.

In cohort 2, after repeated dosing of 10 mg/kg of SNF472, C_max was also observed at the end of the 4‐hour infusion. Concentrations declined rapidly falling below the LLOQ at 5.5 or 8 hours postadministration (Table 2 ).

There was no accumulation in terms of C_max and AUC after repeated dosing of 10 mg/kg.

3.3. PD

In Cohort 1, no significant effect between day of administration was observed, so the results from days 1 and 5 were pooled for each dose and analysed using a 1‐way ANOVA. A statistically significant inhibition of HAP crystallization (p ≤ .001) was observed at all the doses from 3 to 20 mg/kg (approximately 70–80% inhibition). An IC₅₀ of 2.2 mg/kg (133 mg/patient) and an IC₈₀ of 5.6 mg/kg (469 mg/patient) were calculated (Figure 3A). Figure 3B shows the relationship between plasma circulation levels of SNF472 at C_max and the pharmacodynamic effect in all the samples with measurable SNF472 levels (< 0.75 μM). All the doses from 3 to 20 mg/kg increased SNF472 plasma concentrations to levels that were on the pharmacodynamic plateau.

Figure 3.

Intravenous administration of SNF472 inhibits hydroxyapatite crystallization in dialysis patient plasma ex vivo. A, Dose/pharmacodynamics relationship. Four parameter log dose (mg/kg) response modelling of the inhibition of induction of hydroxyapatite crystallization in plasma samples from dialysis patients administered increasing doses of SNF472 or placebo over 5 days. Solid line represents the dose–response curve calculated from the individual data points. Dashed lines represent the 90% confidence interval. B, Pharmacokinetics/pharmacodynamics relationship.

In Cohort 2 (10 mg/kg for 4 weeks), similar inhibition of HAP crystallization was observed at all time points analysed (Figure 4A). The inhibition of crystallization over the entire study period was 78% (Figure 4B).

Figure 4.

Inhibition of hydroxyapatite (HAP) crystallization following repeated dosing of 10 mg/kg SNF472 through the entire study period. A, Time‐course of crystallization inhibition in active patients over the 4‐week period. Solid line connects the means of every day. B, Global inhibition of crystallization throughout the study period. Results represent the mean ± standard deviation. Data were analysed by a Student t test. (*) indicates significant differences vs placebo, p ≤ .001

3.4. Safety

There were no deaths and no TEAEs that led to withdrawal from the study. None of the TEAEs in this study were considered related to the study drug.

In Cohort 1, a total of 3 TEAEs were reported for 2 patients in the SNF472 12.5 mg/kg dose group, and the SNF472 1 mg/kg and 20 mg/kg dose groups each reported 1 TEAE (Table 3). There were no TEAEs in the placebo group or in the SNF472 3 and 5 mg/kg dose groups. No TEAE by preferred term or system organ class was reported for more than 1 patient, and all TEAEs were mild in intensity.

Table 3.

Summary of treatment‐emergent adverse events (TEAEs) in Cohorts 1 and 2

Cohort 1
System organ class Preferred term	Placebo (n = 6) n (%) [#]	SNF472 1 mg/kg (n = 4) n (%) [#]	SNF472 3 mg/kg (n = 4) n (%) [#]	SNF472 5 mg/kg (n = 6) n (%) [#]	SNF472 12.5 mg/kg (n = 6) n (%) [#]	SNF472 20 mg/kg (n = 6) n (%) [#]
Any TEAE	0	1 (25.0) [1]	0	0	2 (33.3) [3]	1 (16.7) [1]
Eye disorders
Ocular hyperaemia	0	0	0	0	0	1 (16.7) [1]
Gastrointestinal disorders
Diarrhoea	0	0	0	0	1 (16.7) [1]	0
Immune system disorders
Hypersensitivity	0	0	0	0	1 (16.7) [1]	0
Nervous system disorders
Headache	0	0	0	0	1 (16.7) [1]	0
Psychiatric disorders
Insomnia	0	1 (25) [1]	0	0	0	0

Cohort 2
System Organ Class Preferred term	Placebo (n = 6) n (%) [#]	SNF472 5 mg/kg (n = 6) n (%) [#]
Any TEAE	1 (50) [1]	2 (33.3) [3]
Gastrointestinal disorders
Diarrhoea	0	1 (16.7) [1]
General disorders and administration site conditions
Pyrexia	0	1 (16.7) [1]
Infections and infestations
Renal cyst infection	0	1 (16.7) [1]
Psychiatric disorders
Transient psychosis	1 (50) [1]	0

In Cohort 2, 3 TEAEs were reported for 2 patients (33.3%) in the study treatment group and 50% (1 of the 2 patients) in the placebo group reported a single TEAE (a serious TEAE of transient psychosis). One of the 6 patients treated with SNF472 10 mg/kg had a serious TEAE (cyst infection). Both serious TEAEs were moderate in intensity.

Physical status, body weight, cardiorespiratory function and body temperature were in the normal range throughout the study, and no clinically relevant effects on heart rate or blood pressure were observed in any of the different TP or in the multiple dose study. Total and ionized calcium levels, phosphate levels and alkaline phosphatase activity were in the normal range, and no effect of SNF472 administration was evidenced (Table 4).

Table 4.

Biochemistry parameters in Cohort 1 subjects

Cohort 1
Parameter	Placebo Mean (CV%) [n]	SNF472 5 mg/kg Mean (CV%) [n]	SNF472 12.5 mg/kg Mean (CV%) [n]	SNF472 20 mg/kg Mean (CV%) [n]
Days 1 and 5
Total calcium (mg/dL)
Predialysis	8.90 (7.51) [12]	9.17 (4.47) [12]	9.18 (4.38) [12]	9.03 (6.04) [10]
Postdialysis	10.0 (9.0) [11]	11.0 (6.6) [12]	11.4 (8.9) [12]	12.1 (12.9) [10]
Ionised calcium (mM)
Predialysis	1.09 (8.04)	1.09 (13.9)	1.16 (5.41)	1.17 (7.00)
Postdialysis	1.26 (6.83)	1.33 (3.66)	1.32 (5.17)	1.33 (3.85)
Phosphate
Predialysis	3.85 (1.84) [2]	3.77 (22.6) [3]	‐	‐
Postdialysis	1.85 (26.8) [2]	1.8 (20.0) [3]	‐	‐
Alkaline phosphatase (IU/L)
Predialysis	63.5 (38.3) [12]	91.3 (81.7) [12]	90.8 (69.8) [12]	93.1 (60.7) [9]
Postdialysis	78.3 (56.8) [12]	90.9 (74.5) [12]	91.4 (64.4) [12]	102 (51) [9]

Five patients had QTcB periods above 500 ms. These changes were not attributable to SNF472 as the effect was seen in placebo and treated. No electrocardiogram changes in any subject were considered to be clinically significant by the Investigator or by an independent cardiology review.

4. DISCUSSION

The objectives of this study were to assess: (1) the safety and tolerability of SNF472 after repeated administration (multiple ascending and fixed repeated doses) in HD patients; (2) the PK characteristics and dialysability after repeated dosing in HD patients; and (3) the activity of different dose levels of SNF472 on the inhibition of the calcification response to an in vitro challenge.

In Cohort 1, ascending doses of placebo, 1, 3, 5, 12.5 and 20 mg/kg were evaluated. After completion of treatment in Cohort 1 and reviewing safety and tolerability, a dose of 10 mg/kg was selected by the DESC for Cohort 2 (repeated administration doses 3 times per week for 4 weeks).

In a previous Phase I study,18 the safety, tolerability and PK of IV SNF472 were evaluated following administration of single IV doses ranging from SNF472 0.5 to 12.5 mg/kg to HV and of a single dose of 9 mg/kg to HD patients. By comparing PK results from equivalent doses of SNF472 in HV and HD patients, the dialysability of the drug was assessed. The PK profile showed a C_max proportionality and a relatively short half‐life. The total amount of SNF472 that was excreted in urine was <1% of the total administered dose. With the exception of mild to moderate local infusion site irritation, SNF472 was well tolerated; more importantly, infusion site irritation was not observed in HD patients and is not expected to be a safety concern for HD patients as SNF472 is diluted prior to entering the circulation.

The current manuscript reports the safety, PK and PD of SNF472 in HD patients after 2 different schemes of treatment, including 5 different doses administered for 1 week (three administrations) and 1 dose administered for 4 weeks (12 administrations). In the current study, the plasma peak concentrations were observed immediately after the end of the 4‐hour infusion (T_max) and declined rapidly to below the LLOQ for all dose levels in both cohorts. C_max and AUC parameters increased in a slightly more than dose‐proportional manner. In both Cohort 1 and Cohort 2, PK parameters showed that there was no accumulation and steady state was not achieved. Dialysability studies show SNF472 to have an insignificant clearance through the dialysis membrane.19 Thus, the absence of clearance due to dialysis, plus the PK parameters, indicate linearity in terms of C_max and AUC and allow a good PK prediction in future clinical studies. In a previous Phase 1 study with HV, it was demonstrated that the total amount of SNF472 excreted in urine accounted for <1% of the total administered dose.16 Since the PK profiles of HV and HD patients are similar,16 the lack of renal function in HD patients is not expected to affect the metabolism of SNF472. Mass balance and quantitative whole‐body autoradiography studies are currently being performed in rats in order to ascertain the complete metabolism and excretion of the compound, but some data on the in vivo metabolism of IP6 in rodents has been already published. Following oral and intravenous administration of radiolabelled IP6 (³H or ¹⁴C) to rats and mice, inositol was the major circulating IP6 metabolite, representing >90% of total circulating radioactivity.20, 21, 22 Lower inositol phosphates appeared quickly in plasma but were not detectable beyond 45 minutes after dosing. An elevated expired ¹⁴CO₂ (about 60% of the administered dose) was found following oral administration of ¹⁴C‐IP6 and ¹⁴C‐inositol to rats,20 and this can be explained by the conversion of inositol into D‐glucuronate and xylitol.23, 24, 25 Therefore, SNF472 is metabolized through sequential dephosphorilation (hydrolysis) and the inositol ring is oxidized to water and CO₂, which is eliminated through expired air.

PD analyses in this study demonstrate that the IV infusion of SNF472 to HD patients at doses from 3 to 20 mg/kg significantly inhibits the induction of HAP crystallization in plasma samples by approximately 70–80% with an IC₅₀ of 2.2 mg/kg (133 mg/patient) and an IC₈₀ of 5.6 mg/kg (469 mg/patient). Repeated administration (3 times per week) of SNF472 at 10 mg/kg for 4 weeks does not affect the crystallization inhibitory potential of the compound. The crystallization inhibitory potential of SNF472 is comparable along all cohorts and is in line with previous PD studies evaluating the effects of SNF472 on HAP crystallization,11 and support the use of SNF472 for the treatment of CVC in HD patients. The design of this Phase 1b trial allowed us to evidence, for the first time, a PK‐PD relationship for SNF472 in HD patients.

IP6, the active ingredient of SNF472, has been shown to prevent CVC in a variety of animal models.12, 13, 26 Published studies12, 13, 26 and nonclinical studies with SNF472 have shown that parenteral IP6 prevents CVC by between 60% and 95% in animal models. As a therapeutic treatment, it could be used at systemic concentrations that would not provoke significant reductions in ionized calcium concentrations due to chelation. The target is the calcium that is deposited on the vessel walls and not the free calcium in solution. The inhibition of CVC by IP6 is accompanied by a reportedly positive effect on bone. Both animal models27 and epidemiological data28, 29, 30 suggest that IP6 could be an effective agent for treating osteoporosis by reducing bone mineral density loss. Animals with an IP6‐enhanced diet reduced the loss of bone mineral density caused by oestrogenic deficiency in ovariectomized Wistar rats.¹⁷

Safety analyses from this study support the safety and tolerability of IV SNF472 in HD patients. There were no deaths and no TEAEs that were considered related to the study drug in either patient cohort. The incidence of TEAEs did not increase with increasing doses of study drug. Electrocardiogram, QTcB intervals and ionized calcium levels were extensively evaluated throughout this study and showed no safety issues associated with treatment with the study drug in the HD patient population. Moreover, disturbance of QT prolongations are frequent findings among patients with chronic kidney disease, especially in ESRD patients on HD.31, 32, 33, 34 A negative correlation has been established between Ca, P and K and QT/QTc interval increase in dialysis.35, 36 Both the QTc interval and the QTc dispersion increases after dialysis,29 in the range of 50–60 ms. Ionized calcium levels were maintained in the normal range, and no clinically relevant effects on heart rate or blood pressure were observed in either patient cohort. It should be noted that the key determinant of serum calcium in HD patients is the calcium concentration of the dialysis fluid. These results complement those obtained in the previous Phase 1 trial in HV and HD patients, now demonstrating that the compound is safe for the dialysis population for a period of 4 weeks. Moreover, the compound does not show accumulation in plasma after 1 month of treatment, while it shows unchanged PD activity after 1 month of treatment and a linear PK/PD relationship.

The results from this study suppose a significant step forward in the development of SNF472 as a novel inhibitor of vascular calcification, by demonstrating the overall safety, tolerability, PK/PD correlation and nondialysability of IV SNF472. This study indicates a positive risk to benefit ratio after treatment for 4 weeks and supports further development of IV SNF472 for the treatment of progression of CVC in HD patients and for the treatment of calcification‐related diseases such as calciphylaxis. Preliminary assessments of the potential for SNF472 to inhibit calcification demonstrated a marked inhibition that is sustained with repeated dosing. Future clinical trials with adequate statistical power need to be conducted to assess the efficacy of IV SNF472 for the treatment/prevention of CVC in HD patients. Two Phase 2 clinical trials with SNF472 are currently ongoing, 1 trial in calciphylaxis patients measuring the effect of SNF472 on ulcer progression, pain and quality of life (NCT02790073, all patients have completed treatment); and 1 trial evaluating the effect of SNF472 treatment for 12 months on CVC progression in ESRD patients (NCT02966028; more than 200 of the planned 400 patients are receiving treatment).

COMPETING INTERESTS

J.P., P.H.J., M.D.F., A.Z.C. and C.S. are employees or have received honoraria from Laboratoris Sanifit SL and are shareholders at Laboratoris Sanifit SL.

CONTRIBUTORS

J.P., P.H.J. and C.S. designed the study; M.D.F., A.Z.C., F.M., V.T., J.M.C. and R.O. supervised the study and carried out experiments; J.P., P.H.J, M.D.F., C.S., F.M. and J.M.C. analysed and interpreted the data; all authors drafted and revised the paper and approved the final version of the manuscript.

Supporting information

Table S1: Summary of patient disposition in Cohort 1 and Cohort 2

Figure S2: Method of SNF472 Administration

Table S3: Concomitant medication of the patients in Cohort 1 and Cohort 2

Salcedo C, Joubert PH, Ferrer MD, et al. A phase 1b randomized, placebo‐controlled clinical trial with SNF472 in haemodialysis patients. Br J Clin Pharmacol. 2019;85:796–806. 10.1111/bcp.13863