Extended-Release Calcifediol Normalized 1,25-Dihydroxyvitamin D and Prevented Progression of Secondary Hyperparathyroidism in Hemodialysis Patients in a Pilot Randomized Clinical Trial Open Access

Serum concentrations of total 25-hydroxyvitamin D (25D) and 1,25-dihydroxyvitamin D (1,25D) decline as chronic kidney disease (CKD) advances [1], becoming insufficient in the absence of effective vitamin D repletion and driving secondary hyperparathyroidism (SHPT). The latest KDIGO clinical practice guideline [2] suggested correcting vitamin D insufficiency (VDI) in non-dialysis (ND) CKD with dietary supplements (either cholecalciferol or ergocalciferol) and using calcitriol (or a 1α-hydroxylated analog) to treat severe and progressive hyperparathyroidism in advanced CKD when 1,25D production is limited by insufficient renal cytochrome P450 25D-1α-hydroxylase (CYP27B1). The guideline did not specify a target for intact parathyroid hormone (iPTH) due to the absence of supporting randomized clinical trial (RCT) data but commented that modest increases may be an appropriate adaptive response to declining estimated glomerular filtration rate. In end-stage kidney disease (ESKD), the guideline suggested maintaining iPTH levels in the range of approximately 2–9 times the upper limit of normal.

Multiple RCTs conducted in ND-CKD patients have established that oral extended-release calcifediol (ERC) effectively raises serum 25D and 1,25D and reduces elevated iPTH despite declining kidney function [3, 4]. Accordingly, ERC is approved as Rayaldee^® in the USA and Europe for treating SHPT in adults with stage 3 or 4 CKD and VDI at doses of 30 µg/day (210 µg/week) escalating, as needed, to 60 µg/day (420 µg/week). These RCTs support the premise that ERC boosts circulating 1,25D by delivering sufficient substrate to extrarenal CYP27B1, consistent with observations that substrate-driven vitamin D hormone production occurs in surgically anephric patients [5, 6]. They further highlight ERC’s potential efficacy in treating advanced SHPT in patients with ESKD requiring dialysis.

The present pilot RCT explored, for the first time, ERC treatment for advanced SHPT in patients undergoing chronic hemodialysis (HD). Specifically, it evaluated the hypothesis that an oral dose regimen of 900 µg/week of ERC could safely and sufficiently raise serum 25D to normalize circulating 1,25D concentrations and, thereby, prevent SHPT progression. The goal of the trial was to inform the design of a subsequent RCT of larger size and longer duration.

This multicenter pilot randomized, single-blind, placebo-controlled trial explored the safety and efficacy of 900 µg/week of oral ERC for raising serum total 25D to ≥50 ng/mL, normalizing circulating 1,25D and reducing plasma iPTH in ESKD patients requiring regular HD. The trial (NCT03602261) was approved by Advarra Institutional Review Board (Columbia, MD; protocol tracking number Pro00025201) and conducted from July 2018 to February 2021. All subjects provided written informed consent prior to trial participation.

Adults (N = 372) requiring HD three times per week for the past 6 months were screened at 22 US dialysis clinics. For inclusion, they must have been receiving vitamin D hormone replacement therapy for at least the preceding month and have plasma iPTH levels of 150 to <600 pg/mL, serum albumin >3.0 g/dL, and serum total 25D <50 ng/mL in the absence of dietary vitamin D supplementation. Approximately half of the screened population were also receiving adjunctive calcimimetic therapy. Exclusion criteria included scheduled kidney transplant or parathyroidectomy and receipt of bisphosphonate therapy or other bone modifying treatment (e.g., denosumab) within the past 6 months. Screen failures (N = 124) resulted most frequently from iPTH or 25D being out of range (73%), unwillingness or inability to comply with trial instructions (10%), inadequate duration of prior SHPT treatment (10%), or other less frequent reasons.

A total of 248 participants met all screening criteria, enrolled, discontinued treatment with calcimimetics, vitamin D hormones and supplements, limited elemental calcium (Ca) intake to ≤1,000 mg per day, and underwent 8 weeks of washout. The first 44 participants who met the post-washout criteria (plasma iPTH increased by at least 50% and into the range of 300 to <1,200 pg/mL, serum 25D <50 ng/mL, corrected serum Ca <9.8 mg/dL, and serum phosphorus (P) < 6.5 mg/dL) were enrolled (from 13 sites) and 163 were not, most frequently because 25D remained ≥50 ng/mL (41%) or iPTH increased insufficiently or was out of range (31%). The minimum post-washout iPTH inclusion criterion of 300 pg/mL was intended to reduce the risk of iPTH oversuppression during the treatment period and the maximum criterion of 1,200 pg/mL was intended to make early terminations due to parathyroidectomy unlikely. Forty-one participants were dismissed prior to randomization after the targeted trial population was achieved.

Eligible participants (N = 44) were randomized 3:1 (in blocks of 4) in single-blind fashion (participants only) according to a statistician-generated code via an Interactive Response System (IRS) to 26 weeks of treatment with ERC (900 µg/week) or placebo, followed by 6 weeks of post-treatment monitoring. The 900 µg weekly dose was selected to be approximately two-fold higher than the maximum weekly dose of 420 µg (60 µg/day) approved for ND-CKD patients in view of the greater severity of SHPT in ESKD. Participants were dosed with two ERC capsules (150 µg each) or two matching placebo capsules at about 1 h into each of 3 weekly HD sessions while monitored by trial staff. Eating was prohibited for 1 h prior to starting dialysis but a low-fat post-dose snack was allowed during dialysis, if requested. Dialysate Ca was maintained at ≤2.5 mEq/L. Appropriate use of phosphate binders was allowed during the treatment period, but use of vitamin D hormone or calcimimetic therapy was prohibited for the duration of the trial. The dose of trial drug was reduced in increments of 2 capsules (300 µg of ERC) per week in the event that plasma iPTH was confirmed (by a second determination obtained at the earliest opportunity) to be <150 pg/mL, corrected serum Ca was confirmed >10.4 mg/dL, or serum P was confirmed >6.5 mg/dL, provided that the elevated serum P was deemed related to the trial drug. Dosing was suspended if plasma iPTH was confirmed to be <100 pg/mL or corrected serum Ca was confirmed >11.0 mg/dL and resumed when plasma iPTH was ≥150 pg/mL and serum Ca was <9.8 mg/dL at a weekly dose reduced by 300 µg or the minimum dosage of 300 µg. Participants were removed from the trial if they experienced more than a 100% increase in iPTH from pretreatment baseline (BL) or exhibited iPTH above 1,200 pg/mL on consecutive visits after 12 weeks of treatment.

Blood samples were collected at the start of the mid- or late-week HD session at BL and at weekly, biweekly, or quarterly intervals depending on the analyte. iPTH, Ca, P, albumin, total 25D, vital signs, pretreatment adverse events (AEs), and treatment-emergent AEs (TEAEs) were monitored at least biweekly. Hematology and full clinical chemistries were monitored at quarterly intervals. Electrocardiograms (ECGs; 12-lead) and brief physical examinations were obtained at BL and the end of treatment (EOT).

Blood samples were shipped to BioReference Laboratories (Elmwood Park, NJ, USA) where serum total 25D was determined as the sum of 25-hydroxyvitamin D₂ and D₃ concentrations by liquid chromatography tandem mass spectrometry (LC-MS/MS), serum total 1,25D by chemiluminescence (Diasorin Liaison) and plasma iPTH by electrochemiluminescence (Roche Elecsys Cobas analyzer with measuring range of 2.4–5,000 pg/mL). Serum samples were forwarded to Syneos Health (QC, Canada) for analysis of calcifediol (25D₃), calcitriol (1,25D₃), and 24,25-dihydroxyvitamin D₃ (24,25D₃) by LC-MS/MS.

A subset of 18 participants from the ERC group were housed in a phase 1 unit for 3 days on two occasions for determination of single-dose or repeated-dose pharmacokinetic (PK) profiles of serum calcifediol and associated pharmacodynamic (PD) profiles of other analytes. A control PK subset of 7 participants from the placebo group were not housed in the phase 1 unit (out of an ethical concern for unnecessary inconvenience to them) but provided blood samples at selected PK timepoints. This procedural difference forced a single-rather than double-blind trial design. Full PK data from these two subsets will be presented elsewhere. Serum calcifediol, calcitriol, 24,25D₃, and total 1,25D were monitored only in the PK subsets at BL and quarterly intervals.

Sample size was based on a prior RCT [3] in ND-CKD patients wherein ERC at doses up to 90 µg/day safely achieved significant iPTH reductions (p < 0.05) versus placebo during an 8-week treatment period. A formal sample size calculation was not undertaken as the data obtained were merely intended to inform the design of a subsequent trial of larger size and longer duration. The intent-to-treat (ITT) population included the 44 participants who were randomized to a treatment group. The safety population included 43 participants who received at least one dose of trial drug. The primary analysis focused on the modified ITT (mITT) population of 37 participants (ERC: N = 28; placebo: N = 9) having at least one 25D and one iPTH measurement recorded at both BL and EOT. The mITT population included 20 participants (ERC: N = 15; placebo: N = 5) from the PK subsets.

BL values for all parameters represented the means of measurements obtained after washout and prior to dosing with trial drug. EOT values represented means of measurements obtained in the last 6 weeks of the 26-week treatment period. Determinations below the limit of quantification (LOQ) were assigned a value equal to the LOQ. BL characteristics were compared between treatment groups using t-tests (for independent samples) for continuous variables and chi-square tests (for categorical variables). Within-group changes from BL to EOT were assessed using t-tests (for paired samples). Between-group comparisons of EOT values were conducted using t-tests (for independent samples). To account for imbalances between treatment groups at BL, analysis of covariance (ANCOVA) was performed with BL values included as covariates. All statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA), and a two-sided p value <0.05 was considered statistically significant. A CONSORT checklist is included as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000546615).

Adults (N = 372) with advanced SHPT and ESKD requiring HD 3 times per week were screened for enrollment in this pilot single-blind (participants only) placebo-controlled RCT. Of these, 248 were enrolled, washed out from prior SHPT therapy and evaluated according to additional post-washout inclusion criteria, yielding 44 eligible participants randomized (3:1) to 26 weeks of treatment with ERC (300 µg per HD; N = 33) or placebo (N = 11). Seven participants were excluded from the mITT population. One withdrew informed consent after randomization and prior to ERC treatment. Four other participants dosed with ERC discontinued prematurely, for reasons unrelated to treatment: withdrawn consent, relocation out-of-state, COVID-19 diagnosis, and physical injury. Two participants dosed with placebo discontinued prematurely: one withdrew consent and the other experienced a confirmed increase in iPTH above the prespecified escape threshold. No participant had a parathyroidectomy or kidney transplant during the trial. Figure 1 displays a consort diagram summarizing the progress of trial participants through the protocol.

Demographic characteristics of the mITT population (N = 37) are summarized in Table 1, in aggregate and by treatment group. Both groups were similar in age, balanced for gender, race, and ethnicity and had comparable body weight and body mass index. All participants had been previously receiving vitamin D hormone replacement therapy and 32% had also been receiving calcimimetic treatment, but further use of these therapies was suspended for the duration of the trial. Clinical characteristics of this population at BL and EOT are summarized in Table 2. The two treatment groups had comparable values at BL for all measured parameters except mean serum 25D, which was 11.9 ng/mL higher (p < 0.01) in the placebo group (36.0 ± 5.3 ng/mL) compared to the ERC group (24.1 ± 1.7 ng/mL).

Mean (±SE) serum 25D rose during ERC treatment to 159.5 ± 10.2 at EOT (p < 0.001) but declined to 30.6 ± 5.5 ng/mL with placebo. Weekly increases averaged 10.0 ± 1.0 ng/mL per week with ERC until a steady-state level of 157.7 ± 10.4 ng/mL was attained after 12 weeks (Fig. 2a), at which time all participants had achieved levels ≥50 ng/mL. Steady-state 25D concentrations in individual participants varied inversely with body weight (R² = 0.1945; p < 0.05; data not shown). Mean 25D levels decreased at a weekly rate of 12.4 ± 1.5 ng/mL during the 6-week follow-up period after ERC dosing had ended. In the PK subgroups (ERC: N = 15; placebo: N = 5), changes in serum levels of calcifediol (25D₃) were consistent with those observed for total 25D.

Serum total 1,25D was below the LOQ (5.0 pg/mL) in 25% of the PK participants at BL and below the lower limit of normal (19.9 pg/mL) in all. Mean 1,25D rose with ERC treatment from 9.4 ± 1.2 pg/mL at BL to 50.7 ± 7.8 at EOT (p < 0.001) but remained unchanged with placebo (12.6 ± 2.7 pg/mL at BL vs. 13.4 ± 4.1 at EOT). Mean plasma iPTH rose 19.8 ± 10.6% with placebo (from 497.6 ± 69.2 pg/mL at BL to 593.1 ± 95.1 at EOT) but decreased 1.7 ± 4.7% with ERC (from 530.4 ± 29.4 pg/mL to 529.6 ± 43.7), the intergroup difference being significant (p < 0.05). Percentage changes in iPTH (from BL) were inversely proportional (R² = 0.2645; p < 0.01) to serum 25D concentrations at EOT (Fig. 2b) with larger decreases occurring primarily at higher 25D exposures.

The intergroup difference in 25D concentrations at BL (24.1 mg/mL in the ERC group vs. 36.0 ng/mL in the placebo group; p < 0.01) may have confounded the observed percent changes in iPTH at EOT. To address this, an ANCOVA was performed with BL 25D included as a covariate in the comparison between groups at EOT. The adjusted least squares means (±SE) showed a 3.7 ± 5.0% iPTH decrease in the ERC group compared to a 26.2 ± 9.4% increase (p = 0.01) in the placebo group.

Nearly all PK participants (93%) treated with ERC exhibited normal 1,25D values by 12 weeks compared to none treated with placebo (p < 0.001). A strong linear correlation was observed between serum total 1,25D and 25D (Fig. 2c) in all serum samples collected from the PK participants at BL and after 12 and 26 weeks of treatment (R² = 0.8248; p < 0.001) which indicated, on average, that 1,25D normalized when 25D was raised to at least 50 ng/mL. Serum 24,25D₃ at BL was below the LOQ (0.54 ng/mL) in all PK participants but became measurable in 92% of the PK participants treated with ERC, rising to 1.7 ± 0.2 ng/mL at EOT (p < 0.001). Changes in serum calcitriol (1,25D₃) were consistent with those observed for total 1,25D.

Mean serum Ca and P remained unchanged in both treatment groups over the 26-week treatment period. Confirmed hypercalcemia was not observed with ERC treatment. One participant in the ERC group required a dose reduction (from 900 to 600 µg/HD) for a confirmed episode of hyperphosphatemia (8.4 mg/dL followed by 8.0) occurring after 22 weeks of treatment. Four participants treated with ERC required dose reductions triggered by confirmed iPTH suppression to levels below 150 pg/mL: three needed a single-dose reduction (from 900 to 600 µg/week) and one required three successive dose reductions (from 900 to 0 µg/week) and then resumed stable dosing at 300 µg. No clinically significant changes were observed in hematology, clinical chemistry, ECGs, vital signs, or physical examinations, and rates of pretreatment AEs and TEAEs were not different between groups. No participants died during the trial.

This prospective placebo-controlled RCT showed, for the first time, that ERC can prevent further progression of advanced SHPT in ESKD patients undergoing regular HD. The trial evaluated the hypothesis that an oral ERC dose regimen (900 µg/week) could safely and sufficiently raise serum total 25D to normalize circulating 1,25D concentrations and, thereby, effectively prevent SHPT progression. Findings from this pilot trial were consistent with the stated hypothesis. Mean serum 25D rose with ERC treatment from 24.1 ± 1.7 ng/mL at BL to a steady-state concentration of 157.7 ± 10.4 after 12 weeks, at which time levels in all participants were above 50 ng/mL. Mean serum total 1,25D rose from 9.4 ± 1.2 pg/mL at BL to 50.7 ± 7.8 at EOT. Nearly all participants (93%) achieved 1,25D values above the lower limit of the normal range (19.9 pg/mL) after 12 weeks. Mean plasma iPTH rose 19.8% with placebo treatment but decreased only 1.7% during ERC treatment (p < 0.05) after the dosage in 4 participants had been reduced for iPTH oversuppression. These percentage changes in iPTH were adjusted to 26% and −4%, respectively, by ANCOVA correction for the intergroup difference in BL 25D concentrations and were inversely proportional to serum 25D exposure. The safety profile of ERC was indistinguishable from placebo throughout the 26-week treatment period: serum Ca, P, and rates of TEAEs were not increased, and hypercalcemia was not observed.

The observed increases in serum 1,25D support the conclusion that ERC can safely restore normal endogenous hormone production and prevent SHPT progression by delivering sufficient substrate to CYP27B1 expressed outside of the kidney. This discovery potentially enables an important and long overdue change to the current paradigm for treating SHPT in ESKD: the seemingly inevitable use of vitamin D hormone therapy and the associated risk of further vascular calcification may be avoided altogether by harnessing substrate-driven 1,25D production in extrarenal tissues expressing CYP27B1 before SHPT becomes severe. Others have also highlighted the importance of early treatment to reduce the risk of medically challenging SHPT [7].

Under the current paradigm, treatment of SHPT typically begins in ND-CKD patients with dietary vitamin D supplements (cholecalciferol or ergocalciferol) in an attempt to correct underlying VDI. VDI is commonly defined as serum 25D levels below either 20 or 30 ng/mL [8‒10]. While intuitively appropriate, these supplements are fat-soluble, prone to accumulation in adipose tissue, and have reduced or unsatisfactory effectiveness in patients who are overweight, as are most who have CKD [11]. The latest KDIGO clinical practice guideline [2] described dietary supplements as “unproven” given that RCT evidence for their effectiveness in raising serum 25D and lowering elevated PTH is lacking in CKD patients [12‒14].

When iPTH concentrations inevitably rise, dietary vitamin D supplements are usually replaced (or augmented) with calcitriol or a 1α-hydroxylated vitamin D analog (e.g., doxercalciferol or paricalcitol) with the rationale that endogenous 1,25D production is impaired due to declining renal CYP27B1 despite KDIGO’s caution against routine use in ND-CKD patients [2]. However, CYP27B1 is expressed outside of the kidney in many different tissues including the parathyroid glands [15] with functionality that has been confirmed in both in vitro and clinical studies. CYP27B1 expressed in cultured human parathyroid tissue has been shown to convert calcifediol to calcitriol [16]. Further, clinical studies have demonstrated that substrate-driven vitamin D hormone production occurs in surgically anephric patients [5, 6]. Similar findings have been reported after administration of calcifediol to ESKD patients with intact residual kidneys [17] as confirmed by the present trial. Extrarenal tissues expressing CYP27B1 require high circulating 25D concentrations to get substrate by passive diffusion rather than by the endocytotic megalin-cubilin transport mechanism found in the kidney [18] which enables delivery of adequate 25D to renal CYP27B1 even when serum 25D levels are low.

Definitive guidance is currently lacking regarding the appropriate target for serum 25D in CKD patients. Vitamin D sufficiency in healthy individuals is inconsistently defined (if at all) as serum total 25D of at least 20 or 30 ng/mL [8‒10]. The latest KDIGO clinical practice guideline [2] concluded that defining a specific target for CKD patients in the current era is likely to be premature. The recently revised Endocrine Society clinical practice guideline for prevention of VDI [19] no longer endorses a serum 25D target of 30 ng/mL to define vitamin D sufficiency or specific 25D levels to define insufficiency and deficiency.

Previous RCTs with ERC indicate that the commonly used targets of 20 or 30 ng/mL for serum 25D are too low [20, 21] in ND-CKD patients with SHPT. One of these RCTs demonstrated that the degree of iPTH reduction was directly proportional to increases in serum 25D above 30 ng/mL [3]. A post hoc analysis of pooled data from two identical and larger RCTs [4] demonstrated that clinically meaningful reductions in iPTH occurred only when serum 25D exceeded at least 50 ng/mL [20]. Findings from the current trial with ERC indicate that restoration of adequate circulating 1,25D concentrations in ESKD patients also requires elevation of serum 25D to levels higher than 50 ng/mL. Together, these trials indicate that a higher 25D target is appropriate in CKD patients with reduced renal capacity to produce 1,25D.

Concerns have been raised by the Institute of Medicine about the safety of raising serum 25D to levels of ≥50 ng/mL [8]. The threshold of toxicity for serum 25D has been postulated by others to be 60–80 [10], 100 [22], 150 [23], or 250 [24, 25] ng/mL, but supportive data are sparse. Data from the present trial show that gradual and sustained elevation of serum 25D to mean (±SE) levels of 157.7 ± 10.4 ng/mL had minimal effects on serum Ca and P concentrations and was free of hypercalcemic episodes and significant TEAEs in HD patients who are highly vulnerable to vitamin D toxicity (due to their inability to excrete excess Ca and P in urine). These data are consistent with previous RCTs demonstrating excellent safety when serum 25D was raised with ERC to much higher levels than 50 ng/mL in ND-CKD patients [3, 4, 20, 26]. They indicate that toxicity may depend more on the rate at which serum 25D is elevated than on the absolute exposure achieved, a concept supported by the lack of toxicity seen in lifeguards experiencing gradual circannual variations of large amplitude [27].

The present trial focused on ESKD patients with advanced SHPT who had previously received vitamin D hormone treatment, often in combination with adjunctive calcimimetic therapy. BL iPTH concentrations were elevated to 522.4 ± 27.5 pg/mL after washout from prior iPTH-lowering therapies and approximately 35% of participants in both treatment groups exhibited iPTH levels exceeding the KDIGO target range of 2–9 times the upper limit of normal, proportions that remained unchanged at EOT. Mean iPTH, Ca and P remained stable during ERC treatment at 300 µg/HD, a dose regimen that produced steady-state 25D concentrations of 157.7 ± 10.4 ng/mL. In contrast, a previous study of 44 HD patients treated with immediate-release calcifediol (IRC) for 17 weeks at an average EOT dose of 364 µg/HD reported increases in serum Ca and P of 1.27 mg/dL (p < 0.001) and 0.49 mg/dL (p < 0.05), respectively, while mean iPTH remained unchanged [28]. Serum 25D levels in that study reached 210–250 ng/mL. Differences in the observed pharmacodynamic profiles of ERC and IRC at these similar doses and 25D exposures resulted, most likely, from the marked differences in the bioavailabilities of ERC (25% [29]) versus IRC (62–77% [30]) and rates of calcifediol release (median time to maximum serum concentration of 21.0 h for ERC [29] versus 4–8 h for IRC [30]). In another study, IRC supplementation for 2 years at 40 µg/HD normalized serum 25D in only 36% of 143 HD patients with VDI (defined as serum 25D <30 ng/mL) without increasing the rates of TEAEs, including hypercalcemia (>10.5 mg/dL) and hyperphosphatemia (>5.5 mg/dL), but changes in iPTH attributable to calcifediol could not be evaluated since coadministration of active vitamin D therapy was allowed [31].

A limitation of the current trial is its small sample size and short duration which rendered it unable to establish the long-term efficacy, safety, and sustainability of ERC’s benefits. A strength of the current trial is the observed consistency between measurements of serum total 25D and 1,25D and parallel measurements of calcifediol (25D₃) and calcitriol (1,25D₃) which alleviates concerns that increases in the total concentrations of these metabolites merely reflected assay interferences. As expected, the levels of the vitamin D₃ metabolites were modestly lower.

In conclusion, in this pilot placebo-controlled RCT, ERC safely raised mean serum 25D, normalized low serum 1,25D, and stabilized elevated iPTH levels in ESKD patients requiring regular HD. The observed increases in serum 1,25D indicated that ERC restored adequate endogenous vitamin D hormone production via substrate-driven conversion to calcitriol in extrarenal tissues expressing CYP27B1, thereby preventing SHPT progression. These findings support a rationale for conducting a subsequent trial of larger size and longer duration.

The authors would like to acknowledge critical reviews of the manuscript by CSL Vifor.

Participants included in the clinical trial described in this article gave written informed consent. The trial (NCT03602261) was conducted in compliance with the principles of the Declaration of Helsinki. The protocol was reviewed by Advarra Institutional Review Board (Columbia, MD; protocol reference no. 00025201).

C.W.B., A.A., J.C., S.A.S., L.L.J., and S.M.S. are or have been employees, consultants, or contractors of OPKO Health, Inc. and have no other conflicts of interest (COI) to disclose. S.M.S. was a member of the journal’s Editorial Board at the time of submission. K.C.N. has no COI to disclose.

This trial was supported by the Renal Division of OPKO Health and CSL Vifor.

C.W.B., A.A., S.A.S., L.L.J., and S.M.S. designed the trial. J.C. and S.A.S. carried out data analysis and preparation of the figures. C.W.B. and S.A.S. drafted the initial version of the manuscript. C.W.B., A.A., J.C., S.A.S., L.L.J., K.C.N., and S.M.S. reviewed, edited, and approved the final version.

The data that support the findings of this trial are not publicly available due to contained information that could compromise the privacy of research participants, but they are available (in redacted form, as applicable) from the corresponding author (C.W.B.) upon reasonable request.