Clinical Kidney Journal, 2021, vol. 14, no. 1, 366–374 doi: 10.1093/ckj/sfz144 Advance Access Publication Date: 30 October 2019 Original Article ORIGINAL ARTICLE Trial design and baseline characteristics of CaLIPSO: a randomized, double-blind placebo-controlled trial of SNF472 in patients receiving haemodialysis with cardiovascular calcification Antonio Bellasi 1, Paolo Raggi2, Jordi Bover3, David A. Bushinsky4, Glenn M. Chertow5, Markus Ketteler6, Mariano Rodriguez7, Smeeta Sinha8, Carolina Salcedo9, Rekha Garg10, Alex Gold10 and Joan Perelló9 1 Research, Innovation and Brand Reputation Unit, ASST Papa Giovanni XXIII, Bergamo, Italy, 2Department of Medicine, Mazankowski Alberta Heart Institute and University of Alberta, Edmonton, AB, Canada, 3 Department of Nephrology, Fundació Puigvert and Universitat Autònoma, IIB Sant Pau, REDinREN, Barcelona, Spain, 4Department of Medicine, University of Rochester Medical Center, Rochester, NY, USA, 5Department of Medicine, Stanford University, Palo Alto, CA, USA, 6Department of General Internal Medicine and Nephrology, Robert-Bosch-Krankenhaus, Stuttgart, Germany, 7Nephrology Unit, Hospital Universitario Reina Sofia, IMIBIC, REDinREN, Córdoba, Spain, 8Department of Renal Medicine, Salford Royal NHS Foundation Trust, Salford, UK, 9 Research and Development, Sanifit Therapeutics, Palma, Spain and 10Research and Development, Sanifit Therapeutics, San Diego, CA, USA Correspondence to: Antonio Bellasi; E-mail: antoniobellasi@gmail.com; Twitter handle: @ABellasi ABSTRACT Background. The objective of CaLIPSO, a Phase 2b, randomized, double-blind, placebo-controlled clinical trial, is to test the hypothesis that myo-inositol hexaphosphate (SNF472) attenuates the progression of cardiovascular calcification in patients receiving maintenance haemodialysis. Here we report the trial design and baseline characteristics of trial participants. Methods. Adult patients on maintenance haemodialysis (6 months) with an Agatston coronary artery calcium score, as measured by a multidetector computed tomography scanner, of 100–3500 U were enrolled. Patients were stratified by Agatston score (100–<400, 400–1000 or >1000 U) and randomized in a 1:1:1 ratio to receive placebo, SNF472 300 mg or SNF472 600 mg administered intravenously three times weekly during each haemodialysis session. Results. Overall, 274 patients were randomized. The mean age of trial participants was 63.6 (standard deviation 8.9) years and 39% were women. The coronary artery, aorta and aortic valve median (25th-75th percentile) Agatston scores at baseline were 730 U (315–1435), 1728 U (625–4978) and 103 U (31–262), respectively, and the median (25th–75th percentile) calcium Received: 22.7.2019; Editorial decision: 13.9.2019 C The Author(s) 2019. Published by Oxford University Press on behalf of ERA-EDTA. V This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/ licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com 366 Trial design and baseline characteristics of CaLIPSO | 367 volume scores at baseline were 666 (310–1234), 1418 (536–4052) and 107 (38–278), respectively. Older age and diabetes mellitus were associated with higher calcium scores at baseline. Conclusions. The CaLIPSO trial enrolled patients on haemodialysis with pre-existent cardiovascular calcification to test the hypothesis that SNF472 attenuates its progression in the coronary arteries, aorta and aortic valve. Keywords: chronic kidney disease, coronary artery calcification, randomized clinical trial, SNF472, vascular calcification INTRODUCTION CaLIPSO is a Phase 2b, multicentre, randomized, double-blind, placebo-controlled clinical trial testing the hypothesis that SNF472 Chronic kidney disease (CKD) is associated with significant mor- compared with placebo attenuates the progression of cardiovascu- bidity, premature mortality, reduced quality of life and high lar calcification in patients receiving maintenance haemodialysis. health care expenditures [1, 2]. Traditional risk factors for the de- In this report we present the design of the CaLIPSO trial and sum- velopment and progression of CKD include advanced age, cardio- marize baseline characteristics of the trial participants. vascular disease, diabetes mellitus, hypertension and obesity [3]. As kidney function declines, a variety of metabolic and hormonal changes occur, many of which result in disordered mineral MATERIALS AND METHODS metabolism, existing within a state of chronic inflammation and multiple organ dysfunction [4–8]. Disorders of mineral metabo- Trial participants lism in CKD include phosphate retention and elevated serum Adult men and women (18–80 years) receiving haemodialysis concentrations of phosphaturic hormones, including parathyroid for 6 months were screened for eligibility. Initially the trial hormone (PTH) and fibroblast growth factor 23 (FGF23). A large recruited patients with an Agatston CAC score between 100 and body of evidence indicates that soft tissue calcifications, 2000 U measured by a multidetector CT scanner (MDCT). An particularly cardiovascular calcification, contribute to cardiovas- amendment to the protocol increased the upper limit to 3500 U, cular morbidity and mortality in patients with CKD [9, 10]. which is more fully representative (and inclusive) of this com- Cardiovascular calcification is an actively regulated process con- plex patient population (Table 1). Ethical approval was obtained trolled by multiple cellular and humoral mechanisms, including in accordance with the local/national processes for each study the loss of inhibitors of mineralization typically expressed by vas- site. All participants provided written informed consent prior to cular cells and the overexpression of osteogenic processes oper- enrolment. The trial, which was conducted according to the ating within vascular lesions, including those that promote principles of the Declaration of Helsinki, was registered at hydroxyapatite nucleation and crystal growth [4]. EudraCT (2016-002834-59) and ClinicalTrials.gov (NCT02966028). The reported prevalence of coronary artery calcification (CAC) increases from 40% in patients with moderate to advanced (Stage 3/4) CKD [11–13] to 60% in patients at haemodialysis initi- Overview of trial design ation and up to 80% in patients receiving maintenance haemo- CaLIPSO consisted of three different phases: screening for dialysis [14]. Cardiovascular calcification may be a modifiable 28 6 3 days, double-blind treatment for 52 weeks and safety as- phenomenon [15]. Thus management guidelines for CKD mineral sessment 4 weeks after administration of the last dose of study and bone disorder (CKD-MBD) recommend the diagnosis and drug (Figure 1). Patients who met eligibility criteria were ran- management of cardiovascular calcification in all patients with domized on Day 1 in a 1:1:1 ratio to receive placebo, SNF472 300 at least moderately decreased kidney function (Stage 3a or higher mg or SNF472 600 mg, three times weekly during haemodialysis CKD) [16]. In these patients, computed tomography (CT) may be sessions. In view of the overall association of Agatston CAC used to determine the presence or absence of cardiovascular cal- score with progression of cardiovascular calcification, randomi- cification; lateral abdominal radiograph for vascular calcification zation was stratified by baseline Agatston CAC score, divided and echocardiogram for valvular calcification are reasonable into three categories: 100–<400, 400–1000 and >1000 U. alternatives [16]. Recommended management strategies to re- Study visits in CaLIPSO were scheduled for screening, base- duce cardiovascular calcification include lowering elevated se- line (Week 1) and Weeks 2, 4, 6, 10, 16, 22, 28, 34, 40, 46, 52 (last rum phosphate while avoiding calcium-based phosphate binders dose) and 56. If the patient terminated study treatment earlier to prevent net positive calcium balance [16]. than planned, then the assessments scheduled for the Week 52 Several large, randomized clinical trials in patients receiving visit were conducted at an early termination visit. dialysis have evaluated a number of therapeutic interventions At screening and Week 52, imaging of the coronary arteries, to reduce cardiovascular events, but they required large sample aorta and aortic valve was performed using MDCT, with a mini- sizes [17–20]. These and other studies did not show a cardiovas- mum of 64 slices. Contiguous 3 mm tomographic slices from cular benefit; thus a significant unmet need remains. above the aortic arch to the diaphragm were obtained at end- SNF472 is an intravenous formulation of the hexasodium expiration and reconstructed during diastole or at the time of least salt of myo-inositol hexaphosphate (IP6, phytate), a compound motion. All scans were centrally reviewed and each calcified area that targets vascular calcification by binding to the growth sites was quantified using both volume [23] and Agatston score [24]. of the hydroxyapatite crystal and inhibiting hydroxyapatite nu- The Agatston score is more sensitive to the calcium content of a cleation and crystal growth [21, 22]. IP6 is a naturally occurring plaque, but it is less reproducible than the volume score [25], thus substance found in beans, brown rice, corn, sesame seeds, the calcium volume score was selected as the primary endpoint. wheat bran and other high-fibre foods; however, oral absorption Total hip and femoral neck bone mineral density (BMD) were is low. IP6 is normally present in minimal concentrations (in the measured at screening and Week 52 by dual-energy X-ray ab- micromolar range) in body fluids [21, 22]. sorptiometry. Imaging modality, anatomical positioning, 368 | A. Bellasi et al. Table 1. Trial eligibility criteria Inclusion criteria Female or male patients CAC score of 100–3500 U inclusive within a 4-week period prior to randomization, as measured by an MDCT scanner Patients 18–54 years of age at randomization with a history of diabetes mellitus or 55–80 years of age at randomization (with or without a his- tory of diabetes mellitus) Patients receiving haemodialysis for 6 months prior to randomization Exclusion criteria Scheduled date for kidney transplant from a known living donor Weight >300 lb (136 kg) Hospitalization in the previous 3 months prior to randomization for unstable angina, myocardial infarction, stroke, transient ischeamic at- tack, amputation or peripheral or coronary bypass surgery History of unstable heart failure in the previous 3 months, defined as an unplanned presentation to a hospital or dialysis treatment facility with signs/symptoms of acute pulmonary oedema and requiring ultrafiltration therapy History of cancer that has been in remission for <5 years prior to randomization. A history of basal cell carcinoma or Stage 1 squamous cell carcinoma of the skin is allowed Pregnant or trying to become pregnant, currently breastfeeding or of childbearing potential (including perimenopausal women who have had a menstrual period within 1 year) and not willing to practice birth control using a double-barrier method (criteria apply to women only) at least 30 days after the last dose of study medication Hypocalcaemia, defined as serum calcium <8.0 mg/dL (or 2.0 mmol/L) for the serum calcium most proximal to screening per patient’s medi- cal records Extreme elevation in serum phosphorous, defined as phosphorous >10 mg/dL (or 3.23 mmol/L) within the last 2 months proximal to screen- ing per patient’s medical records Uncontrolled hypertension, defined as any two or more consecutive post-dialysis diastolic blood pressure measurements >100 mmHg within the last 2 months proximal to screening Expected survival <2 years in the investigator’s medical opinion Known active drug or alcohol abuse within 1 year of randomization Use of other investigational drugs within 30 days of randomization Non-compliance with dialysis treatment that, in the opinion of the investigator, as evidenced by either repeated missed dialysis treatments or significant non-compliance with the patient’s medication regimen Inability to comply with all required study procedures and schedules, inability to speak and read in the protocol-derived language of that patient’s clinical site or unwillingness or inability to give written informed consent FIGURE 1: CaLIPSO trial flow chart. In Step 1, potential study participants who satisfied the inclusion and exclusion criteria underwent an assessment by MDCT scanner to determine the Agatston score for the coronary artery, as well as dual-energy X-ray absorptiometry for BMD of the total hip and femoral neck. Patients with con- firmed calcification of the coronary artery (initially 100–2000 U; later 100–3500 U) at Step 1 entered Step 2 to complete all other screening assessments and confirm all el- igibility criteria were met. coverage and parameter were consistent across both imaging showing minimal or no lesser trochanter on the scanned image. visits for each patient. The left femur was scanned at both vis- For the Hologic scanner (Marlborough, MA, USA), the array its; if a fracture, prosthesis, deformity or problem prohibited mode was used for each image, while for the Prodigy Lunar scanning of the left femur, then the right femur was scanned at scanner (GE Healthcare, Chicago, IL, USA) automatically chose both visits. A foot positioner was used to keep the femoral shaft thin (<13 cm), standard (13–25 cm) or thick (>25 cm) depending straight and parallel to the edge of the scanned image, with the on the body thickness and height/weight of the patient. greater trochanter centred vertically in the window. The entire Samples for pharmacokinetic and biomarker analyses were femoral head was visible, with a 25 internal rotation of the hip collected from a subset of participants at selected sites before Trial design and baseline characteristics of CaLIPSO | 369 and after the infusion of SNF472/placebo at baseline and Weeks proportional hazards (Cox) regression is used to compare 10, 22 and 52. Laboratory analyses, including haematology, co- SNF472 to placebo, stratified by baseline Agatston CAC score. agulation and chemistry, were performed at baseline and The planned sample size for CaLIPSO was approximately 270 Weeks 10, 28, 40 and 52 via a central laboratory. patients. Based on available data [26], assuming a standard de- A Data and Safety Monitoring Board (DSMB) was formed to viation (SD) of the mean change in CAC volume score of 0.30 on monitor patient safety and data integrity during the study. In the log scale (from baseline to Week 52) and a dropout rate of January 2019, a futility analysis was performed based on data 25%, the sample size (i.e. 90 patients per group) was selected to from the first 136 patients and the DSMB recommended contin- provide 80% power to test the hypothesis that the log- uation of the study. transformed true difference in progression between the com- bined SNF472 doses and the placebo group is 0.126, correspond- ing to a true ratio of 1.134, or 19% progression for the average of Statistical analysis the SNF472 doses and 35% progression for the placebo group. In this report we describe baseline characteristics of patients enrolled in the study, including demographics, medical history, RESULTS current and prior medications and calcium scores (Agatston and calcium volume). We calculated the Spearman’s correlation Enrolment between the Agatston CAC score and CAC calcium volume score Of 645 patients from 72 centres who were screened for enrol- at baseline. We used an exploratory multiple linear regression ment, 239 did not meet the inclusion criteria, most of whom model to identify associations with the log-transformed CAC vol- (n ¼ 231) did not have an Agatston CAC score within the re- ume score at baseline. Based on prior research [26], candidate quired range at screening. Other reasons for not entering the variables in the initial model were sex (male/female), age (years), randomization are shown in Table 2. A total of 274 patients dialysis vintage (months), PTH, magnesium, diabetes (yes/no), were randomized to one of the three treatment groups. Of these, non-calcium-based phosphate-binder use (yes/no), statin use 82 (30%) had an Agatston CAC score of 100–399 U, 77 (28%) had a (yes/no), warfarin use (yes/no) and history of overt atheroscle- score of 400–1000 U and 115 (42%) had a score >1000 U. rotic cardiovascular disease (yes/no). Stepwise regression was used to identify factors associated with baseline log-transformed CAC volume score with 0.10 significance level for entry into the Demographics and clinical and laboratory model and 0.10 significance level for remaining in the model. characteristics at baseline Analyses were conducted using SAS version 9.3 or above (SAS The mean age of the 274 randomized patients was 63.6 years Institute, Cary, NC, USA). Results for body mass index were not (SD 8.9) and 39% were women (Table 3). Demographic character- included in this baseline analysis. istics and vital signs were similar across the Agatston CAC score The primary endpoint of the study will be the change in log cor- strata. Most patients (93%) had hypertension, 62% had diabetes onary artery calcium volume score from baseline to Week 52 for mellitus, 22% had clinical history of coronary artery disease and the combined dose groups versus placebo. Secondary endpoints 16% had congestive heart failure. Patients with higher Agatston are change from baseline in log Agatston CAC score at Week 52; CAC scores had a numerically higher prevalence of diabetes proportion of patients with <15% progression in Agatston CAC mellitus, coronary artery disease and congestive heart failure score at Week 52; change in thoracic aorta calcium score from than those with lower Agatston CAC scores. The median dura- baseline to Week 52; change in aortic valve calcium score from tion of haemodialysis was 42.4 months. baseline to Week 52; incidence of composite safety endpoint that Laboratory parameters were generally similar across the includes death from cardiovascular causes, myocardial infarction, Agatston CAC score strata (Table 4). In each treatment group, stroke or heart failure; mortality rate (all-cause and cardiovascular); serum PTH was well above the normal range (10–65 pg/mL) and change from baseline in levels of selected biomarkers including C- magnesium was slightly above the normal range (1.7–2.2 mg/ reactive protein, fetuin-A, FGF23, matrix gla protein (MGP), sclero- dL). However, serum PTH concentrations were consistent with stin and growth differentiation factor 15; change in BMD between Kidney Disease: Improving Global Outcomes (KDIGO) recom- baseline and 52 weeks and safety. Exploratory endpoints include mendations for patients with kidney failure [16]. Serum calcium change from baseline in pulse pressure, systolic blood pressure and albumin were in the normal range in each treatment group. and diastolic blood pressure at Weeks 28 and 52 in all patients. Approximately one-third of the patients were treated with a The modified intention-to-treat analysis population is used calcium-based phosphate binder and half were treated with the for all primary and secondary efficacy analyses. This population non-calcium-based phosphate binder, sevelamer; 7% were consists of all randomized patients who receive at least one dose of SNF472 or placebo and have at least one post- Table 2. Summary of reasons for screening failure randomization evaluable CT scan with a CAC score. The safety population will include all randomized patients who receive at Reason patient did not meet entry criteria All screened (N ¼ 645) least one dose of SNF472 or placebo. The CAC volume score is log-transformed prior to analysis. The primary analysis model Inclusion criteria 239 Exclusion criteria 30 is an analysis of covariance (ANCOVA) with the change in log Both inclusion and exclusion 1 score (log[Week 52] – log[baseline]) as the dependent variable, Withdrew consent 42 and a fixed effects term for the randomized treatment group as Lost to follow-up 0 well as log(baseline) as a covariate; the model is stratified by Other 59 baseline Agatston CAC score. For secondary efficacy outcomes, CT not completed/not evaluable 19 ANCOVA is used to compare combined SNF472 doses versus Kidney transplant 2 placebo and SNF472 300- and 600-mg doses versus placebo indi- Screening/enrolment closed 32 vidually. For the event-driven secondary outcomes, Kaplan– Other (reason not specified) 6 Meier product limit estimates are used to display the data and 370 | A. Bellasi et al. Table 3. Demographic and baseline characteristics Coronary artery calcium score (Agatston) category 100–399 U 400–1000 U >1000 U Total Characteristic (n ¼ 82) (n ¼ 77) (n ¼ 115) (N ¼ 274) Age (years), mean 6 SD 61.0 6 9.1 62.9 6 8.7 65.9 6 8.3 63.6 6 8.9 Sex, n (%) Male 51 (62) 42 (55) 74 (64) 167 (61) Female 31 (38) 35 (45) 41 (36) 107 (39) Race, n (%) White 54 (66) 50 (65) 84 (73) 188 (69) Black or African American 21 (26) 19 (25) 21 (18) 61 (22) Asian 4 (5) 4 (5) 2 (2) 10 (4) Other 0 0 4 (4) 4 (1) Not reported 3 (4) 4 (5) 5 (4) 12 (4) Ethnicity, n (%) Hispanic or Latino 31 (38) 29 (38) 40 (35) 100 (36) Body mass index (kg/m2) 28.4 6 6.7 29.6 6 6.2 29.0 6 5.6 29.0 6 6.1 Vital signs, mean 6 SD Systolic blood pressure (mmHg) 138.8 6 21.5 136.1 6 25.6 136.4 6 26.6 137.0 6 24.9 Diastolic blood pressure (mmHg) 72.3 6 11.6 70.1 6 13.6 65.5 6 13.6 68.8 6 13.3 Medical history, n (%) Diabetes mellitus 48 (59) 47 (61) 75 (65) 170 (62) Hypertension 77 (94) 71 (92) 107 (93) 255 (93) Peripheral vascular disease 13 (16) 9 (12) 14 (12) 36 (13) Cerebrovascular accident 11 (13) 6 (8) 15 (13) 32 (12) Myocardial infarction 9 (11) 6 (8) 16 (14) 31 (11) Coronary artery disease 12 (15) 15 (19) 34 (30) 61 (22) Congestive heart failure 10 (12) 12 (16) 21 (18) 43 (16) Time on haemodialysis (months) Median (25th–75th percentile) 37.9 (19.1–66.1) 44.3 (19.5–77.0) 43.4 (20.4–74.8) 42.4 (19.5–74.8) <12 months, n (%) 9 (11) 7 (9) 15 (13) 31 (11) 12–36 months, n (%) 31 (38) 26 (34) 36 (31) 93 (34) >36 months, n (%) 42 (51) 44 (57) 64 (56) 150 (55) Table 4. Baseline mineral metabolism parameters Coronary artery calcium score (Agatston) category 100–399 U 400–1000 U >1000 U Total Baseline laboratory parameter (N ¼ 82) (n ¼ 77) (n ¼ 115) (N ¼ 274) Parathyroid hormone (pg/mL), median (25th–75th percentile) 312 (240–482) 318 (213–501) 362 (238–587) 335 (231–530) Magnesium (mg/dL), mean 6 SD 2.5 6 0.4 2.5 6 0.4 2.4 6 0.4 2.5 6 0.4 Albumin (g/dL), mean 6 SD 4.0 6 0.3 4.0 6 0.3 3.9 6 0.3 4.0 6 0.3 Uncorrected serum calcium (mg/dL), mean 6 SD 8.7 6 0.8 9.0 6 0.8 8.9 6 0.6 8.8 6 0.7 Corrected serum calciuma (mg/dL), mean 6 SD 8.7 6 0.7 9.0 6 0.7 8.9 6 0.7 8.9 6 0.7 Alkaline phosphatase (U/L), mean 6 SD 110.8 6 48.7 112.6 6 62.0 118.6 6 81.8 114.7 6 68.1 a Corrected serum calcium ¼ uncorrected serum calcium þ 0.8 (4  serum albumin). treated with lanthanum and 12% with another non-calcium- (25th–75th percentile) and geometric means (95% confidence in- based phosphate binder (Table 5). Statins were used by 62% of terval). Median Agatston scores and calcium volume scores, re- patients, calcimimetics by 28%, warfarin by 7% and activated vi- spectively, were 730 U and 666 for the coronary arteries, 1728 U tamin D by 53%. More patients with the highest Agatston CAC and 1418 for the aorta and 103 U and 107 for the aortic valve scores at baseline (>1000 U) were being treated with non- (Table 6 and Figure 2). For all patients combined, the correlation calcium-based phosphate binders, calcimimetics, statins or at baseline between Agatston CAC scores and CAC calcium vol- warfarin, whereas patients with lower Agatston CAC scores ume scores was 0.968 (P < 0.001). As expected, median Agatston were more likely to be treated with calcium-based binders. scores and calcium volume scores in the aorta and the aortic valve were numerically higher in patients with a baseline Measures of cardiovascular calcification Agatston CAC score >1000 U than in those with a baseline score of 100–399 U. Among the 115 patients with an Agatston CAC Baseline calcium scores (Agatston and volume) for coronary ar- score >1000 U at baseline, the score at the aorta and aortic arch tery, aorta and aortic valve were not normally distributed. Thus was >100 U for 110 (96%) and 44 (38%) patients, respectively. summary statistics at each location are presented as medians Trial design and baseline characteristics of CaLIPSO | 371 Table 5. Baseline medication use Coronary artery calcium score (Agatston) category 100–399 U 400–1000 U >1000 U Total Baseline medication (n ¼ 82) (n ¼ 77) (n ¼ 115) (N ¼ 274) Calcium-based phosphate binders, n (%) 27 (33) 20 (26) 35 (30) 82 (30) Non-calcium-based phosphate binders, n (%) Sevelamer 37 (45) 39 (51) 61 (53) 137 (50) Lanthanum 4 (5) 3 (4) 12 (10) 19 (7) Other 10 (12) 6 (8) 18 (16) 34 (12) Calcimimetics, n (%) 20 (24) 21 (27) 37 (32) 78 (28) Statins, n (%) 47 (57) 44 (57) 78 (68) 169 (62) Warfarin, n (%) 3 (4) 3 (4) 12 (10) 18 (7) Activated vitamin D, n (%) 47 (57) 39 (51) 58 (50) 144 (53) Table 6. Baseline calcification assessments using Agatston and volume methods Agatston calcium score Calcium volume score Location and baseline Median Geometric Median Geometric Agatston score (25th–75th percentile) mean 6 SE (25th–75th percentile) mean 6 SE Coronary artery 100–399 U (n ¼ 82) 210 (149–285) 207 6 9 204 (163–278) 212 6 11 400–1000 U (n ¼ 77) 598 (479–757) 602 6 19 596 (441–692) 564 6 17 >1000 U (n ¼ 115) 1617 (1222–1903) 1600 6 52 1378 (1068–1640) 1324 6 44 Total (n ¼ 274) 730 (315–1435) 659 6 45 666 (310–1234) 602 6 37 Aorta 100–399 U (n ¼ 77) 984 (427–2429) 926 6 465 807 (344–2096) 830 6 368 400–1000 U (n ¼ 72) 1446 (411–3904) 1102 6 570 1217 (360–3040) 969 6 440 >1000 U (n ¼ 113) 3620 (1305–7302) 2617 6 785 2837 (1099–6063) 2219 6 605 Total (n ¼ 262) 1728 (625–4978) 1520 6 407 1418 (536–4052) 1323 6 315 Aortic valve 100–399 U (n ¼ 42) 73 (31–192) 60 6 31 55 (37–209) 69 6 27 400–1000 U (n ¼ 41) 76 (31–231) 89 6 46 86 (46–219) 92 6 37 >1000 U (n ¼ 74) 137 (32–395) 102 6 73 136 (37–353) 113 6 57 Total (n ¼ 157) 103 (31–262) 85 6 38 107 (38–278) 94 6 30 SE, standard error. FIGURE 2: Distribution of baseline coronary artery calcium scores (Agatston and calcium volume). 372 | A. Bellasi et al. Table 7. Linear regression analysis of predictors of log-transformed coronary artery calcium volume scores at baseline Variable Parameter estimate Standard error P-value Model R2 Intercept 4.875 0.409 <0.001 – Age (years) 0.022 0.006 <0.001 0.039 Coexistent diabetes 0.255 0.111 0.022 0.060 Candidate variables in the initial model were sex (male/female), age (years), dialysis vintage (months), PTH, magnesium, diabetes (yes/no), non-calcium-based phos- phate binder use (yes/no), statin use (yes/no), warfarin use (yes/no) and history of overt atherosclerotic cardiovascular disease (yes/no). Stepwise regression was used to identify factors associated with baseline log-transformed CAC volume score with 0.10 significance level for entry into the model and 0.10 significance level for remaining in the model. Factors associated with CAC score at baseline cinacalcet with placebo on top of standard care, showed gener- ally favourable effects on cardiovascular events and fracture, In the multivariable-adjusted linear regression analysis that re- but the primary endpoint was also not met [42]. quired a 0.10 significance level for variables to remain in the The pathogenesis of cardiovascular calcification in CKD is model, older age (P < 0.001) and diabetes mellitus (P ¼ 0.022) complex and only partly understood [10]. Irrespective of the were independently associated with higher baseline CAC triggering factors, calcium–phosphate crystals may deposit in volume scores (Table 7). the soft tissues if there is an imbalance of calcification inhibi- tors and promoters [43]. In the presence of both high and low DISCUSSION mineral turnover, it has been postulated that bone cannot Cardiovascular calcification is highly prevalent in patients with accommodate additional calcium and phosphate in the matrix, impaired kidney function; at least 80% of patients receiving leading to excess circulating ions and eventually precipitation maintenance haemodialysis exhibit some degree of cardiovas- of these minerals in soft tissue [44]. Cardiovascular calcification cular calcification [14, 27]. Cardiovascular calcification is also has often been described as a passive degenerative process due associated with adverse outcomes [28–31]. Among 181 consecu- to ‘wear and tear’ and progressive senescence [5, 45]. However, tive patients with non-dialysis-requiring CKD, baseline CAC active cellular mechanisms are principally responsible for ec- score and CAC score progression predicted the occurrence of topic calcification [6, 10, 45]. Laboratory experiments show that cardiac death and/or myocardial infarction [32]. Numerous excess calcium and/or phosphate, along with decreased levels observations support the contribution of valvular and vascular of calcification inhibitors, induce vascular smooth muscle cells calcification to morbidity and mortality in CKD [33]. CAC has to differentiate and assume an osteoblastic phenotype. These been associated with increased arterial stiffness [34, 35], left cells initiate deposition of collagen, which becomes calcified in ventricular hypertrophy [36], epicardial adipose tissue thickness the arterial wall [43]. Micronutrients such as magnesium and and volume, myocardial ischaemia [37] and pro-arrhythmic phosphorus may have a similar role in valvular calcification electrocardiographic abnormalities [35]. Valvular calcification [45]. Further stimulus towards the accumulation of soft tissue causes restriction of leaflet motion [38, 39], increased transval- calcification in uraemia is provided by the frequent depletion of vular gradient [38, 39] and left atrium enlargement [38, 39] and inhibitors of calcification [46]. For example, low levels of fetuin-A, is associated with reduced survival [40]. osteoprotegerin, MGP, vitamin K and pyrophosphate have been In this study, patients with higher Agatston CAC scores at consistently described in patients with CKD and associated with baseline were more likely to be treated with a non-calcium- vascular calcification [10]. based phosphate binder, calcimimetics, statin or warfarin. To date, there are no approved therapies to directly slow the These management strategies were consistent with recent progression of cardiovascular calcification. Although a number KDIGO CKD-MBD guidelines [16]. Intense research has been un- of therapies potentially target the calcification process, such as dertaken in recent years to identify modifiable risk factors that pyrophosphate, bisphosphonates, thiosulphate and vitamin K, induce calcification, as well as drugs that may reduce its pro- none of these compounds have been shown to be effective in or gression, in the hope of reducing the incidence of cardiovascu- approved for the treatment or prevention of cardiovascular lar events and mortality in patients with impaired kidney calcification. Current guidelines consider patients with CKD and function [16]. Multiple prior studies have tested the effects of a cardiovascular calcification to be at the highest cardiovascular variety of interventions that modify mineral metabolism on cal- risk and suggest managing the biochemical abnormalities cification of the coronary arteries, aorta and cardiac valves. observed in patients with CKD as a therapeutic strategy to Treatment with non-calcium-based phosphate binders has mitigate progression of cardiovascular calcification [16]. been shown to attenuate CAC progression compared with Preclinical data showed that SNF472 reduced development calcium-based phosphate binders [41]. Similarly, in a random- of cardiovascular calcification by up to 80% [47]. In early clinical ized clinical trial (ADVANCE), cinacalcet was compared with trials, SNF472 was well tolerated in humans [48, 49]. The placebo on top of standard care including calcitriol or vitamin D CaLIPSO study was designed to test the hypothesis that SNF472 analogues in patients receiving haemodialysis with moderate- slows the progression of calcification in the coronary arteries, to-severe secondary hyperparathyroidism. Cinacalcet was asso- aorta and aortic valve. CaLIPSO was powered to detect a clini- ciated with slower progression of calcification from baseline to cally meaningful reduction in CAC as measured by CAC volume Week 52 in the thoracic aorta (cinacalcet 19%, control 33%; scores between the combined SNF472 dose groups versus pla- P ¼ 0.055), aortic valve (cinacalcet 6%, control 52%; P ¼ 0.014) and cebo after 52 weeks of treatment. mitral valve (cinacalcet 12%, control 54%; P ¼ 0.053), but the trial The CaLIPSO study has limitations. This Phase 2 trial was did not meet its primary endpoint for mean change in Agatston not powered to evaluate a reduction in cardiovascular mortality CAC score (cinacalcet 24%, control 31%; P ¼ 0.073) [26]. EVOLVE, and morbidity. Future appropriately powered studies will be a large event-driven randomized clinical trial comparing needed to assess the effects of SNF472 on cardiovascular Trial design and baseline characteristics of CaLIPSO | 373 outcomes. The relatively short follow-up (52 weeks) and the use REFERENCES of different CT scanners at study sites are other potential limita- 1. Thompson S, James M, Wiebe N. Cause of death in patients tions. To increase the quality and comparability of evaluations, with reduced kidney function. J Am Soc Nephrol 2015; 26: a common procedure for CT scanning acquisition was used and 2504–2511 CT scans were evaluated in a central core laboratory by readers 2. Losito A, Nunzi E, Pittavini L et al. Cardiovascular morbidity blinded to treatment assignment using common methods for and long term mortality associated with in hospital small calcium quantitation in the coronary arteries, including increases of serum creatinine. J Nephrol 2018; 31: 71–77 Agatston and volume scores. Although multiple factors may af- 3. Tsai WC, Wu HY, Peng YS et al. Risk factors for development fect the progression of cardiovascular calcification, patients and progression of chronic kidney disease: a systematic re- recruited in the CaLIPSO study were only stratified according to view and exploratory meta-analysis. Medicine (Baltimore) baseline Agatston CAC score, a powerful predictor of CAC pro- 2016; 95: e3013 gression. Serum phosphate, which was collected in a subset of 4. Moranne O, Froissart M, Rossert J et al. Timing of onset of patients, was not available for this analysis. Data on dialysis CKD-related metabolic complications. J Am Soc Nephrol 2009; session duration and dialysate calcium concentration were not 20: 164–171 collected. 5. Kooman JP, Kotanko P, Schols AM et al. Chronic kidney dis- In summary, the CaLIPSO trial was designed to investigate ease and premature ageing. Nat Rev Nephrol 2014; 10: 732–742 the inhibitory effect of SNF472 on the progression of cardiovas- 6. Feldreich T, Nowak C, Fall T et al. Circulating proteins as pre- cular calcification in patients on maintenance haemodialysis. dictors of cardiovascular mortality in end-stage renal dis- The results will determine further evaluation of SNF472 in man- ease. J Nephrol 2019; 32: 111–119 aging cardiovascular disease in this patient population. 7. Quinones H, Hamdi T, Sakhaee K et al. Control of metabolic predisposition to cardiovascular complications of chronic FUNDING kidney disease by effervescent calcium magnesium citrate: a feasibility study. J Nephrol 2019; 32: 93–100 The study is sponsored by Sanifit Therapeutics. Clinipace per- 8. Yamamoto S, Fukagawa M. Uremic toxicity and bone in CKD. formed the statistical analyses with funding from Sanifit J Nephrol 2017; 30: 623–627 Therapeutics S.A. Medical writing support from Jonathan 9. Bellasi A, Raggi P. Vascular imaging in chronic kidney dis- Latham (PharmaScribe) was funded by Sanifit Therapeutics S.A. ease. Curr Opin Nephrol Hypertens 2012; 21: 382–388 10. Henaut L, Chillon JM, Kamel S et al. Updates on the mecha- nisms and the care of cardiovascular calcification in chronic AUTHORS’ CONTRIBUTIONS kidney disease. Semin Nephrol 2018; 38: 233–250 All authors contributed to the research idea and study design. 11. Kramer H, Toto R, Peshock R et al. Association between M.R. and S.S. contributed to the acquisition of data. All authors chronic kidney disease and coronary artery calcification: the contributed to the analysis and interpretation of data. Each au- Dallas Heart Study. J Am Soc Nephrol 2005; 16: 507–513 thor contributed important intellectual content during manu- 12. Matsushita K, Sang Y, Ballew SH et al. Subclinical atheroscle- script drafting or revision and accepts accountability for the rosis measures for cardiovascular prediction in CKD. J Am overall work by ensuring that questions pertaining to the accu- Soc Nephrol 2015; 26: 439–447 racy and integrity of any portion of the work are appropriately 13. Gorriz JL, Molina P, Cerveron MJ et al. Vascular calcification investigated and resolved. A.B. takes responsibility that this in patients with nondialysis CKD over 3 years. Clin J Am Soc study has been reported honestly, accurately and transparently; Nephrol 2015; 10: 654–666 that no important aspects of the study have been omitted and 14. Raggi P, Boulay A, Chasan-Taber S et al. Cardiac calcification that any discrepancies from the study as planned and registered in adult hemodialysis patients. A link between end-stage re- have been explained. nal disease and cardiovascular disease? J Am Coll Cardiol 2002; 39: 695–701 15. Bover J, Evenepoel P, Urena-Torres P et al. Pro: cardiovascular CONFLICT OF INTEREST STATEMENT calcifications are clinically relevant. Nephrol Dial Transplant A.B. received personal fees from Sanifit. P.R. received other 2015; 30: 345–351 compensation from Sanifit. J.B. received personal fees from 16. Kidney Disease: Improving Global Outcomes CKD-MBD Sanifit, Sanofi-Genzyme, Vifor-Fresenius-Renal Pharma, Update Work Group. KDIGO 2017 clinical practice guideline AbbVie, Amgen and Shire. D.A.B. received personal fees from update for the diagnosis, evaluation, prevention, and treat- Sanifit, Tricida, Relypsa/Vifor/Fresenius, Sanofi/Genzyme and ment of chronic kidney disease–mineral and bone disorder Amgen and grants from the National Institutes of Health and (CKD-MBD). Kidney Int Suppl 2017; 7: 1–59 Renal Research Institute. G.M.C. received personal fees from 17. Wanner C, Krane V, Marz W et al. Atorvastatin in patients Akebia, AMAG, Amgen, Ardelyx, AstraZeneca, Gilead, Reata, with type 2 diabetes mellitus undergoing hemodialysis. N Sanifit and Vertex; has an ownership interest in Ardelyx, Engl J Med 2005; 353: 238–248 CloudCath, Cricket, Durect, Outset, PuraCath and Physiowave 18. Wheeler DC, London GM, Parfrey PS et al. Effects of cinacalcet and received research funding from Amgen and Janssen. M.K. on atherosclerotic and nonatherosclerotic cardiovascular received personal fees from Sanifit, Amgen, Medice, Sanofi and events in patients receiving hemodialysis: the EValuation Of Vifor. M.R. received personal fees from Amgen, Sanofi, Vifor, cinacalcet HCl therapy to Lower cardioVascular Events and Sanifit. S.S. received personal fees from Sanifit, Vifor (EVOLVE) trial. J Am Heart Assoc 2014; 3: e001363 Fresenius and Napp. C.S. and J.P. are employees and sharehold- 19. St Peter WL, Liu J, Weinhandl E et al. A comparison of seve- ers of Sanifit Therapeutics S.A., and have patents related to lamer and calcium-based phosphate binders on mortality, SNF472. R.G. and A.G. are employees and shareholders of Sanifit hospitalization, and morbidity in hemodialysis: a secondary Therapeutics S.A. analysis of the Dialysis Clinical Outcomes Revisited (DCOR) 374 | A. Bellasi et al. randomized trial using claims data. Am J Kidney Dis 2008; 51: cardiac repolarization deterioration in hemodialysis 445–454 patients. Kidney Blood Press Res 2011; 34: 180–187 20. Eknoyan G, Beck GJ, Cheung AK et al. Effect of dialysis dose 36. Kitamura K, Fujii H, Nakai K et al. Relationship between and membrane flux in maintenance hemodialysis. N Engl J cardiac calcification and left ventricular hypertrophy in Med 2002; 347: 2010–2019 patients with chronic kidney disease at hemodialysis initia- 21. Perelló J, Gómez M, Ferrer MD et al. SNF472, a novel inhibitor tion. Heart Vessels 2017; 32: 1109–1116 of vascular calcification, could be administered during he- 37. Karohl C, D’Marco L, Bellasi A et al. Hybrid myocardial imag- modialysis to attain potentially therapeutic phytate levels. ing for risk stratification prior to kidney transplantation: J Nephrol 2018; 31: 287–296 added value of coronary calcium and epicardial adipose tis- 22. Shears SB. Assessing the omnipotence of inositol hexaki- sue. J Nucl Cardiol 2013; 20: 1013–1020 sphosphate. Cell Signal 2001; 13: 151–158 38. Movva R, Murthy K, Romero-Corral A et al. Calcification of 23. Callister TQ, Cooil B, Raya SP et al. Coronary artery disease: the mitral valve and annulus: systematic evaluation of improved reproducibility of calcium scoring with an effects on valve anatomy and function. J Am Soc Echocardiogr electron-beam CT volumetric method. Radiology 1998; 208: 2013; 26: 1135–1142 807–814 39. Pressman GS, Movva R, Topilsky Y et al. Mitral annular dy- 24. Agatston AS, Janowitz WR, Hildner FJ et al. Quantification of namics in mitral annular calcification: a three-dimensional coronary artery calcium using ultrafast computed tomogra- imaging study. J Am Soc Echocardiogr 2015; 28: 786–794 phy. J Am Coll Cardiol 1990; 15: 827–832 40. Raggi P, Bellasi A, Gamboa C et al. All-cause mortality in 25. Achenbach S, Ropers D, Mohlenkamp S et al. Variability of re- hemodialysis patients with heart valve calcification. Clin J peated coronary artery calcium measurements by electron Am Soc Nephrol 2011; 6: 1990–1995 beam tomography. Am J Cardiol 2001; 87: 210–213 41. Patel L, Bernard LM, Elder GJ. Sevelamer versus calcium- 26. Raggi P, Chertow GM, Torres PU et al. The ADVANCE study: a based binders for treatment of hyperphosphatemia in CKD: randomized study to evaluate the effects of cinacalcet plus a meta-analysis of randomized controlled trials. Clin J Am low-dose vitamin D on vascular calcification in patients on Soc Nephrol 2016; 11: 232–244 hemodialysis. Nephrol Dial Transplant 2011; 26: 1327–1339 42. Chertow GM, Block GA, Correa-Rotter R et al. Effect of 27. Bellasi A, Ferramosca E, Muntner P et al. Correlation of sim- cinacalcet on cardiovascular disease in patients undergoing ple imaging tests and coronary artery calcium measured by dialysis. N Engl J Med 2012; 367: 2482–2494 computed tomography in hemodialysis patients. Kidney Int 43. Raggi P, Giachelli C, Bellasi A. Interaction of vascular 2006; 70: 1623–1628 and bone disease in patients with normal renal function 28. Bellasi A, Block GA, Ferramosca E et al. Integration of clinical and patients undergoing dialysis. Nat Rev Cardiol 2007; 4: and imaging data to predict death in hemodialysis patients. 26–33 Hemodial Int 2013; 17:12–18 44. Bellasi A, Lacey C, Taylor AJ et al. Comparison of prognostic 29. Block GA, Raggi P, Bellasi A et al. Mortality effect of coronary usefulness of coronary artery calcium in men versus women calcification and phosphate binder choice in incident hemo- (results from a meta- and pooled analysis estimating dialysis patients. Kidney Int 2007; 71: 438–441 all-cause mortality and coronary heart disease death or 30. Raggi P, Cooil B, Shaw LJ et al. Progression of coronary cal- myocardial infarction). Am J Cardiol 2007; 100: 409–414 cium on serial electron beam tomographic scanning is 45. Bellasi A, Di Lullo L, Raggi P. Cardiovascular calcification: the greater in patients with future myocardial infarction. Am J emerging role of micronutrients. Atherosclerosis 2018; 273: Cardiol 2003; 92: 827–829 119–121 31. Budoff MJ, Hokanson JE, Nasir K et al. Progression of coronary 46. Leopold JA. Vascular calcification: mechanisms of vascular artery calcium predicts all-cause mortality. JACC Cardiovasc smooth muscle cell calcification. Trends Cardiovasc Med 2015; Imaging 2010; 3: 1229–1236 25: 267–274 32. Russo D, Corrao S, Battaglia Y et al. Progression of coronary ar- 47. Ferrer MD, Ketteler M, Tur F et al. Characterization of SNF472 tery calcification and cardiac events in patients with chronic pharmacokinetics and efficacy in uremic and non-uremic renal disease not receiving dialysis. Kidney Int 2011; 80: 112–118 rats models of cardiovascular calcification. PLoS One 2018; 13: 33. Brown AJ, Teng Z, Evans PC et al. Role of biomechanical e0197061 forces in the natural history of coronary atherosclerosis. Nat 48. Perelló J, Joubert PH, Ferrer MD et al. First-time-in-human Rev Cardiol 2016; 13: 210–220 randomized clinical trial in healthy volunteers and haemo- 34. Raggi P, Bellasi A, Ferramosca E et al. Association of pulse dialysis patients with SNF472, a novel inhibitor of vascular wave velocity with vascular and valvular calcification in he- calcification. Br J Clin Pharmacol 2018; 84: 2867–2876 modialysis patients. Kidney Int 2007; 71: 802–807 49. Salcedo C, Joubert PH, Ferrer MD et al. A phase 1b random- 35. Di Iorio B, Nargi O, Cucciniello E et al. Coronary artery calcifi- ized, placebo-controlled clinical trial with SNF472 in haemo- cation progression is associated with arterial stiffness and dialysis patients. Br J Clin Pharmacol 2019; 85: 796–806