Abstract and Introduction
Purpose of Review: Vascular calcification (VC) is associated with increased cardiovascular event rates, particularly in patients with end-stage kidney disease (ESKD). Dysregulated mineral metabolism and inflammation have been shown to promote VC, however, treatment options targeting VC specifically are not available. This review outlines the pathophysiological mechanisms contributing to VC in ESKD and describes recent studies evaluating the effects of the first-in-class inhibitor of VC, SNF472.
Recent Findings: SNF472 directly inhibits calcium phosphate crystal formation and aggregation. SNF472 has completed early phase clinical trials with a favourable safety profile and Phase 2 clinical trial data have shown attenuation of coronary artery and aortic valve calcification in patients receiving hemodialysis.
Summary: Therapeutic agents that directly target VC may prevent the multiple complications associated with dystrophic calcification in patients with ESKD.
Vascular calcification (VC) increases with ageing and its presence is associated with higher rates of cardiovascular event in the general population. Patients with chronic kidney disease (CKD) have accelerated VC which starts in the early stages of CKD. Moreover, the presence of VC predicts poor outcomes. Data from the Chronic Renal Insufficiency Cohort showed that a 1 SD log increase in coronary artery calcification (CAC) score was associated with a 40% increased risk of cardiovascular disease, 44% increased risk of myocardial infarction and a 39% increased risk of heart failure. Patients with end-stage kidney disease (ESKD) have a particularly high burden of cardiovascular disease and risk of cardiovascular mortality; 85% of incident dialysis patients over the age of 45 have evidence of cardiovascular disease.
Therapeutic agents that improve cardiovascular outcomes in the general population have failed to improve cardiovascular outcomes in patients receiving maintenance hemodialysis (HD).[5,6] Given that VC is associated with worse cardiovascular outcomes in patients with ESKD, it is plausible that interrupting or reversing this process would be of benefit. This review will outline the pathophysiological mechanisms contributing to VC in ESKD and outline studies that have attempted to evaluate the effects of the first-in-class inhibitor of VC, SNF472.
Pathophysiology of Vascular Calcification
The pathophysiology of VC in CKD is well described in a comprehensive review by Moe et al.. In brief, VC is multifaceted and includes an active cell-mediated process as well as the physiochemical deposition of calcium and phosphate. VC in ESKD is propagated by an imbalance between inducers and inhibitors of calcification within the vascular compartment. Patients receiving HD have dysregulated calcium and phosphate metabolism as well as chronic inflammation, all of which have been shown to promote VC. The imbalance is further exacerbated by a reduction in naturally occurring inhibitors of calcification such as matrix-Gla protein, pyrophosphate and fetuin-A. This creates a pro-calcific environment that promotes the osteogenic transformation of vascular smooth muscle cells (VSMC) and hydroxyapatite (HAP) formation.
Calciprotein particles (CPP) are circulating complexes of calcium, phosphate and proteins; they have been associated with the development of VC. Primary CPP develop into secondary CPP which contain HAP crystals and are a risk factor for the development of VC. This highly regulated cell-mediated process results in the deposition of a mineralised matrix similar to that seen in bone formation. The final common pathway involves the formation and growth of HAP crystals (Figure 1). Patients with advanced CKD and VC have been reported to have larger circulating CPP than those without VC. HAP crystal formation and propagation is the final common pathway in the development of VC, whether intimal or medial.
Multiple factors are implicated in the development of vascular calcification. SNF472 physiochemically blocks the formation of calcium phosphate crystals and thereby inhibits the formation of new hydroxyapatite crystals. Ca2+, calcium; PO4 3-, Phosphate; FGF23, fibroblast growth factor 23; HAP, hydroxyapatite. Image reproduced with permission from Sanifit Therapeutics.
Therapeutic agents that are able to prevent HAP formation may attenuate the progression of VC and its potentially deleterious consequences of cardiovascular disease and death.
SNF472 Mechanisms of Action
Myo-inositol hexaphosphate (IP6, phytate) is a naturally occurring substance found in foods with high fibre content such as cereal and brown rice but has poor oral bioavailability. Myoinositol hexaphosphate is found in humans at low concentrations and has been shown to prevent calcification in vitro and in vivo. In humans, urinary phytate levels have been shown to correlate with dietary phytate intake. Fernandez-Palomeque et al. reported that patients with cardiac valve calcification had lower urinary phytate levels relative to patients with no valvular calcification, suggesting a potential protective role for phytate. Patients on maintenance HD may also be phytate deficient as phytate is highly water soluble and therefore dialyzable. Patients are also often placed on phosphate restricted diets which invariably result in a reduction in dietary phytate. These factors may further contribute to the VC seen in patients receiving HD.
SNF472 is the hexasodium salt of myo-inositol hexaphosphate and is being developed as an intravenously administered inhibitor of VC which can be administered via HD lines directly into the bloodstream over 2.5–3 h.
SNF472 is a potent and rapid inhibitor of in vitro calcium phosphate crystal formation. In vitro studies confirm that SNF472 binds to the surface of an early HAP nucleus or on to the face of a growing HAP crystal and prevents further crystal formation and aggregation. SNF472 has also been shown to inhibit VC in nonuremic rats by 60–70% and by 80% in adenine-induced uremic rats. Importantly, SNF472 prevented the progression of VC in vitamin D-treated rats with preexisting VC.
The relation between calcium phosphate crystallization and progression of calcification has been studied in vivo and in clinical trials. In rats treated with vitamin D to induce VC, SNF472 dose-dependently inhibited the formation of calcium phosphate crystals which in turn correlated with inhibition of VC.
Perello et al. further delineated the mechanism of action of SNF472 in a series of studies to evaluate the affinity of SNF472 to HAP, inhibition of HAP crystal formation and inhibition of HAP crystallisation. SNF472 binding to HAP is concentration dependent and is saturated at approximately 7 μM and SNF472 remains bound for at least 7 days. SNF472 also inhibited HAP crystallization in a concentration-dependent manner without a significant increase in apoptosis of VSMC despite exposure to SNF472 for 5 days. The authors investigated the effect of SNF472 in rat VSMCs that had been incubated with Ca3 (PO4)2 to initiate calcification. SNF472 at a concentration of 1 μM inhibited calcification by 67%. There was also evidence of inhibition of osteogenic differentiation; SNF472 inhibited Ca3(PO4)2-induced down-regulation of the smooth muscle cell marker alpha smooth muscle actin and prevented up-regulation of markers of osteogenic differentiation (alkaline phosphatase and Cbfa1). These data suggest that SNF472 may modulate osteogenic differentiation of VSMC as well as prevent crystallisation of HAP. Given that SNF472 inhibits formation of HAP, studies were also undertaken to evaluate the effects of SNF472 on bone mineralization. Dogs treated with SNF472 for 9 months did not exhibit any differences in bone histomorphology or the osteoclast biomarker, tartrate-resistant acid phosphatase, compared to control animals. These data show that SNF472, at doses being evaluated for clinical study, inhibit the final common pathway in VC.
Current treatment strategies for aortic stenosis involve valve replacement or valvuloplasty (although the latter fails to result in sustained improvement in valve function); there are no pharmacological interventions. SNF472 has been shown to inhibit calcification of aortic valve interstitial cells. SNF472 dose-dependently inhibited experimentally induced in vitro human aortic valve calcification using valvular interstitial cells (VICs). Healthy heart valves were collected from heart transplant recipients and calcified aortic valve leaflets were collected during aortic valve replacement. The inhibitory effect was more pronounced in VICs cultured from calcified valves compared to healthy valves. SNF472 was also able to inhibit calcification even after one week in osteogenic medium; calcification was inhibited by 81% and 100% (P < 0.01) at concentrations of 30 μM and 100 μM, respectively. These findings suggest that SNF472 has the potential to attenuate progression of VC in tissues with preexisting calcification. These preclinical studies highlight the potential for SNF472 to be evaluated as a modifier of calcific aortic valve stenosis, the prevalence of which is increasing in our aging dialysis population.
Peripheral Arterial Disease in End-stage Kidney Disease
Peripheral arterial disease (PAD) is highly prevalent in patients with ESKD compared to the general population with reported prevalence of between 18.6% and 37.8%.[18–20] A large observational study interrogated the German Diagnosis-Related Groups reimbursement system and established that 2.7 fold more patients with PAD-ESKD had severe disease as assessed by the Rutherford Classification compared to patients with PAD and normal or near normal kidney function (49.7% vs 18.5%). Patients with PAD-ESKD were more likely to undergo amputation and also had significantly higher mortality after 4 years (78% vs 27%). In keeping with other cardiovascular disorders in CKD, nontraditional risk factors are likely to be a significant driver in the development of PAD seen in patients with ESKD (PAD-ESKD). Medial calcification is thought to play a major role in PAD-ESKD, particularly in the distal vessels of the foot and the tibial arteries. Medial VC, associated reductions in perfusion associated with vascular stiffness, and other factors unrelated to the 'usual culprits' (i.e., diabetes, hypertension, hypercholesterolemia, and tobacco use) are likely responsible for particularly poor outcomes after revascularization in PAD-ESKD relative to the general population. SNF472 therefore also provides a potential therapeutic intervention in this troublesome entity given its effects on VC.
The effects of SNF472 on lower limb blood perfusion, VC and walking ability were evaluated in a rat model of PAD-ESKD. Rats were treated with vitamin D for 3 days to induce calcification and limb ischemia. After 5 days, rats received no treatment or daily subcutaneous SNF472, subcutaneous saline (SNF472 vehicle control), oral cilostazol or oral cilostazol vehicle control. Rats were sacrificed at day 14 and aortic and femoral artery calcium content was quantified. Walking ability was evaluated using a treadmill test at baseline, day 4 and day 11. Laser doppler imaging was used to measure posterior limb blood perfusion at baseline, day 4 and day 13. Rats treated with SNF472 were the only group that demonstrated a significant improvement across all three parameters. Mean walking distance increased by 49% compared to the saline vehicle control. Aortic and femoral artery calcification was reduced by 41% and 31%, respectively, in comparison to saline vehicle-treated rats. Cilostazol, which one might expect to have beneficial effects in a traditional PAD model, had no significant effect on any of the parameters. These data support the hypothesis that PAD-ESKD has different etiological factors than traditional PAD and may be amenable to treatments that target VC.
SNF472 Clinical Studies
Phase 1 clinical studies evaluating SNF472 in healthy volunteers and patients receiving HD patients have yielded a favourable side effect profile; a total of 19 healthy volunteers and 218 patients have received SNF472 in Phase 1 and 2 clinical trials. The Phase 1 studies included single-dose and multiple-dose studies which exposed 21 HD patients to either 9 mg/kg, 1–20 mg/kg for 1 week, or 10 mg/kg for 4 weeks.[25,26]
Subsequent Phase 2 clinical trials have been undertaken in calcific uremic arteriolopathy ('calciphylaxis') and cardio VC in patients receiving maintenance HD.
This devastating disorder of skin arteriolar calcification is associated with distressingly high rates of morbidity and mortality. The effect of SNF472 on wound healing and pain in patients with calciphylaxis was evaluated in a phase 2 open-label single arm clinical trial. Fourteen patients receiving thrice weekly HD and with a clinical diagnosis of calciphylaxis received approximately 7 mg/kg SNF472 during each HD session over 2.5–3 h. Wound healing, pain scores and health-related quality of life significantly improved over the course of the 12-week treatment period. Spectrometric measurements of calcium phosphate crystallization from the serum of the patients in this study showed consistent inhibition of crystallization by almost 70% over the course of 12 weeks. The open label nature of the trial makes it difficult to draw firm conclusions but results were promising enough to embark on a Phase 3 international placebo-controlled randomised clinical trial. This prospective placebo-controlled randomized clinical trial will recruit 66 patients to evaluate the effect of SNF472 in patients receiving thrice weekly dialysis with a confirmed clinical diagnosis of ulcerated calciphylaxis. A central wound rating group will confirm the diagnosis before 1:1 randomization to 7 mg/kg SNF472 or placebo at on each HD session or placebo. After 12 weeks the study will enter a 12-week open label phase where all subjects will receive thrice weekly SNF472 during dialysis. Primary endpoints are changes in wound healing and pain score.
The CaLIPSO (Cal for calcium and ipso meaning the item itself) study was a phase 2b international, multicentre, randomized, double-blind, placebo-controlled clinical trial which recruited 274 patients with preexisting VC as measured by baseline CAC scores. The primary endpoint was change in log CAC volume score from baseline to week 52. Inclusion criteria included a CAC Agatston score between 100 and 3500 units. Adult patients between the ages of 18 and 54 also required a history of type l or type 2 diabetes. Baseline Agatston score was used to stratify patients (100–399, 400–1000 or >1000 Units). Patients were randomised 1:1:1 to receiver placebo, 300 mg SNF472 or 600 mg SNF472. Study drug or placebo was administered thrice weekly during HD over 2.5 ± 0.5 h. Baseline characteristics were well matched across the 3 treatment groups. It is important to note that patients in CaLIPSO were receiving treatments recommended by KDIGO and which are thought to confer a benefit for the development of VC, notably noncalcium based binders and calcimimetics. The study met its primary endpoint with the combined dosing groups of SNF472 showing a reduction in the progression of mean CAC volume score of 11% (95% CI, 7–15%) compared with 20% (95% CI, 14–26%) in the placebo group (P < 0.016). A mean change of 12% (95% CI, 6–18%) was reported for the 300 mg dose group and a mean change of 10% (95% CI, 4–17%) was reported for the 600 mg dose group. Similar results were obtained using the CAC Agatston score. The attenuation in aortic valve calcification was larger in magnitude and in keeping with the preclinical data; the mean change in calcium volume score in the combined SNF472 groups was 14% [95% CI, 5–24%] compared with 98% [95% CI, 77–123%] in the placebo group (P < 0.001). The change in thoracic aorta calcium volume score was numerically lower in the SNF472 combined groups, but the difference was not statistically significant (23% [95% CI, 16–30%] vs. 28% [95% CI, 19–38%], P = 0.40). Baseline thoracic calcium volume score was high in all patients which may have made it difficult to detect a clinically significant difference in a vascular bed that was already heavily calcified. Whether SNF472 is able to attenuate earlier aortic calcification requires further study.
Pharmacodynamic assays showed that calcium phosphate crystallization was inhibited by at least 50% in patients who received SNF472; moreover, there was a dose-dependent effect on crystallization: 15% ± 17% for placebo, 61% ± 19% for SNF472 300 mg, and 75% ± 9% for SNF472 600 mg (mean ± SD). The results of the PD assay also correlated with the primary efficacy endpoint of CAC volume score, with slower rates of progression in patients with higher levels of calcium phosphate crystallization. These data lend further support that inhibition of HAP formation can translate into clinically meaningful reductions in vascular and heart valve calcification in patients receiving HD.
SNF472 inhibits formation of HAP and as a result the Phase 2b CaLIPSO trial also explored the effect of SNF472 on bone mineral density (BMD) using dual-energy X-ray absorptiometry (DXA). Of the 274 patients randomized in CaLIPSO, 202 had baseline and evaluable follow-up DXA. The correlation between BMD and fracture in patients on HD has not been established; however, there appears to be a more consistent inverse relation between VC and BMD. There were no differences in serum albumin, intact PTH, calcium, or phosphate between the placebo and SNF472 groups but the majority of patients in both groups experienced a modest decline in BMD during the course of the study (52 weeks). The reduction in BMD at the femoral neck was numerically larger in patients treated with SNF472. Clinical fractures were infrequent in both groups. Analysis of safety data and adverse events did not identify any clinically significant abnormalities in clinical laboratory values and there were no safety signals identified by the Drug and Safety Monitoring Board.
The presence of VC is associated with age, diabetes dialysis vintage, preexisting atherosclerotic and cardiovascular disease as well as commonly prescribed medications including statins, warfarin, phosphate binders and calcimimetics. Predefined subgroup analyses demonstrated that SNF472 had similar efficacy across all these groups. The consistency of results across all subgroups supports the hypothesis that SNF472 is able to attenuate VC via the final common pathway irrespective of potential etiological factors.
Curr Opin Nephrol Hypertens. 2021;30(4):424-429. © 2021 Lippincott Williams & Wilkins