Additional Proxy Soliciting Materials (Form DEFA14A)

Approximately 37 million US adults have CKD, a major public health concern, as it is a progressive condition that culminates in end-stage kidney disease (ESKD).^1,2 The economic/societal costs of CKD are

substantial.^2,3 In 2018, ESKD represented 7.2% US Medicare spending.⁴ CKD is underreported with inconsistency in kidney disease testing.⁵ Large registry-based studies indicate that D-CKD is the most common cause of ESKD.⁶

Correspondence: Guido Filler, University of Western Ontario, 800 Commissioners Road East, E3-206, London, Ontario N6A 5W9, Canada. E-mail:[email protected]

Received 26 March 2022; accepted 11 April 2022

The pathomorphologic sequence of nephron loss with glomerular decapitation and progressive tubular fibrosis has been described in diabetic nephropathy.⁷ Once nephron loss begins, no new organoids can be formed.⁷ There is no cure for CKD and treatments are

small molecules targeting biochemical pathways in the kidney to affect related comorbidities; however, the underlying glomerular and tubulointerstitial dysfunction remains unaltered. Cell-based therapies are a promising treatment, and in preclinical CKD models, they reduce inflammation and fibrosis, resulting in kidney function stabilization.^8,9

In animal models, we revealed that isolated and expanded SRCs injected into diseased kidneys augment kidney function and improve survival.^10,11 On the basis of this evidence, we are conducting Federal Drug Administration-approved human clinical trials.^12-14 Elucidating the mechanism of how SRC injections restore kidney function in humans is key to understanding the potential of this therapy for stabilizing CKD.

SRCs are the bioactive product in REACT derived from the patients' own kidney cells. The REACT product is an admixture principally of epithelial cells from the proximal tubules and glomeruli and smaller numbers of other cell subpopulations, such as interstitial cells, collectively called SRCs.¹⁵ Animal models suggest that kidney restoration with SRC-based therapies may mirror fetal kidney development.¹¹ REACT may afford the diseased kidney components of the developmental pathway to initiate a cascade of events that halt disease progression and/or restore kidney function.^16,17 We present early findings in a subpopulation of 22 patients with moderate to advanced type 2 D-CKD who received REACT as a part of our larger ongoing clinical trial (NCT02836574). We describe change in kidney function and characterize the progenitor cell lines of the REACT product with FACS analysis of membrane-bound nuclear transcription factors, representing cap mesenchyme, ureteric bud, and glomerular lineages (Supplementary Appendix 1), correlating levels of these factors with observed kidney outcomes. We also evaluated secretion of VEGF-A in conditioned media sourced from human SRCs. This is a proof-of-concept clinical trial suggesting that SRC therapy may enable kidney function stabilization through neo kidney-like tissue.¹⁶

Study Population

We enrolled 30- to 80-year-old patients with type 2 D-CKD from multiple institutions across the USA, who had an eGFR of 20 to 50 ml/min per 1.73 m² and had consented to the parent study.¹⁸ The diagnosis of type 2 D-CKD was made by clinical diagnosis and did not require histopathologic evidence. Their comorbidities were managed with standard of care, and sodium-glucose transport protein 2 receptor blockade

treatment was not a contraindication. Except for diabetes-related conditions, patients with incapacitating comorbidities were excluded. Our cohort only included selected REACT-treated patients from the parent study who consented to this publication and whose SRCs were characterized by FACS.

Biopsy and REACT Injection Procedures

All patients underwent an image-guided standard percutaneous kidney biopsy to isolate and expand kidney cells, creating autologous homologous SRCs. The SRCs were formulated in thermolabile hydrogel to manufacture a fresh REACT product. The first computed tomography (CT)-guided percutaneous injection of patient's autologous SRCs occurred approximately 3 months after the biopsy, and the second percutaneous injection occurred 7 ± 3 months after the first injection (based on clinical observations and trial protocols).^18,19 There were 2 patients who received only 1 injection. The REACT dosing was patient specific, based on kidney volume calculated by magnetic resonance imaging. The cumulative REACT dose ranged from 8 to 16 ml (8.0-16 × 10⁸ SRCs) for 20 patients receiving 2 injections and 4.5 to 5.5 ml (4.5-5.5 × 10⁸ SRCs) for 2 patients who received only 1 injection. All procedures were under conscious sedation protocols using i.v. midazolam and fentanyl and same-day-admission outpatient visits.

Clinical and Laboratory Data

Clinical and laboratory data were collected at the time of the first REACT injection and every 3 months for 1 year after the first injection. Biochemical measures included electrolytes Hgb, sCr, urea nitrogen, phosphorus, calcium, potassium, bicarbonate, glycated Hgb, log-transformed intact parathyroid hormone, and log-transformed urinary albumin/sCr ratio (UACR). eGFR was estimated using the 2009 CKD-EPI formula for IDMS traceable serum sCr alone and the 2012 CKD-EPI formula for sCr and cystatin C (nephelometry with certified reference materials).^20,21 We analyzed the cohort before and after they received the first REACT injection, establishing their eGFR slope pre- and post-intervention. Patients were classified based on their annualized eGFR slopes (in ml/min per 1.73 m² per year) as low responders (<0), moderate responders ($0 and ≤2), and high responders (>2), but for analysis purposes, we combined the moderate and high responder groups and compared them with low responders.

Serious Adverse Events

Serious adverse events (SAEs) were determined by event seriousness and intensity per reporting from the

Medical Dictionary for Regulatory Activities version 23.0 into Preferred Terms and System Organ Classes.

Ethics and Trial Registration

The trial was approved by the research ethics board of the participating centers and registered at http:// clinicaltrials.gov/ct2/show/NCT02836574. The first patient consented to the study on March 9, 2017. The trial protocol was approved by each site's Institutional Review Board or Ethics Committee on human research and the patients provided written informed consent to perform cellular analyses of their SRCs.

Phenotypic Marker Analysis of SRCs

FACS analysis determined whether the SRCs contained markers associated with cap mesenchyme, ureteric bud, and glomeruli. The analysis of these proteins, which are mostly transcription factors to describe SRC phenotype (Supplementary Appendix 1), was per-formed as described by Burnette and Bruce in 2013.²²

VEGF-AEnzyme-Linked Immunosorbent Assay

Three different human SRC cultures (TCHK-030, 031, 032) produced identically to the REACT product from cadaveric donor tissue with conditioned media, and negative control media were analyzed. Individual and average VEGF concentrations were measured at 1-month intervals in triplicate using the Human VEGF Quantikine enzyme-linked immunosorbent assay (R&D Systems), according to the manufacturer's protocol.

Statistics

We used simple descriptive statistics for screening demographic and laboratory measures. Normally distributed parameters (determined by Shapiro-Wilk normality test) were expressed as mean ± SD. Not normally distributed parameters were log-transformed. Longitudinal data were compared using a longitudinal linear mixed effect models with a correlated random intercept and slope (for time).²³ Patient identifier was a random effect, time was a fixed effect, and eGFR was the dependent variable. This analysis was summarized as the annualized slope: the average change in eGFR over 1 year time. To perform this analysis, the pre-injection annualized slope used all measurements from screening until and including day of injection, as this measurement was taken before the injection. The postinjection annualized slope used all measurements from the day after injection to the end of each patient's follow-up. All retest measurements and unscheduled visits were included in the analysis for both the pre-and post-injection analyses. We did not perform a power calculation, as the sample size was small. Some patients had a 24-hour UACR rather than a random sample, and here the random UACR was imputed.

Patients

A total of 28 patients were eligible for this interim analysis. Six did not consent to have their data published. In the final sample of 22 patients, 2 received only 1 injection of REACT; after their first injection, 1 patient commenced on clopidogrel and the second one had uncontrolled hypertension. In the remaining 20 patients with 2 REACT injections, blood pressure did not change (data not shown). Because only 3 patients were on phosphate binders, 2 on erythropoietin supplementation, and 4 on potassium binders, we did not analyze the impact of REACT on phosphorus, Hgb, or potassium levels. Of the 22 patients, 10 were treated with an angiotensin-converting enzyme inhibitor, 13 patients received an angiotensin II receptor blocker, and 3 patients were treated with sodium-glucose transport protein 2 inhibitors.

Laboratory values at screening are depicted in Table 1, with a mean CKD-EPI 2009 eGFR of 37.3 ± 8.91 ml/min per 1.73 m² and a mean glycated Hgb of 7.0 ± 1.05%. Consistent with the expected decline, the eGFR dropped to 33.0 ± 8.91 ml/min per 1.73 m² at the time of the first REACT injection. Using linear mixed effect model, the annualized eGFR slope improved significantly between pre-injection (-3.98 ml/min per 1.73 m² per year) and post-injection of REACT (-1.27 ml/min per 1.73 m² per year) in the full cohort (P = 0.032).

Using the 2012 CKD-EPI equation based on both sCR and cystatin C, the annualized eGFR slope improved from -4.63 ml/min per 1.73 m² per year preinjection

BMI, body mass index; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; sCR, serum creatinine.

Figure 1. Response of eGFR (creatinine and cystatin C, 2012 CKD-EPI) to SRC injection. Although the overall group had a stabilization of the eGFR decline, 7 patients had a sustained increase of their eGFR. Day 0 represents the date of the first injection. Using linear mixed effects model, the annualized eGFR slope significantly improved from -4.63 ml/min per 1.73 m² per year to -1.69 ml/min per 1.73 m² per year (P = 0.015). CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; REACT, Renal Autologous Cell Therapy; SRC, selected renal cell.

Table 2. Linear mixed effects model analysis of various clinical parameters comparing preinjection and postinjection annualized slope

BUN, bound urea nitrogen; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; PTH, parathyroid hormone; sCR, serum creatinine;UACR, urinary albumin/sCr ratio.

Table 3. Serious adverse events by system organ class and preferred terminology based on the medical dictionary for regulatory activities. N = 15 of 22 patients

eGFR over time changes depended on the expression of RET on the REACT product. This interaction trended toward significance (P = 0.09).

Serious Adverse Events

SAEs were common in this population due to the comorbidities of advanced diabetic kidney disease (DKD) and metabolic syndrome but were similar to other historical CKD trials. No SAEs were associated with the biopsies and REACT injections. A total of 61 SAEs occurred in 15 of 22 patients, with the 5 most common categories being cardiac, infectious, renal, respiratory, and metabolic (see Table 3). Among those with 2 REACT injections, 2 had fatal SAEs and 1 progressed to ESKD. One patient experienced squamous cell carcinoma of the lung, causing left lung collapse/postobstructive pneumonia and hypercalcemia, resulting in death 12 months post second REACT injection. An autopsy was declined by the family. A second patient had a myocardial infarction, and the third patient (screening eGFR of 33 ml/min per 1.73 m² and heavy proteinuria) had rapid decline in eGFR to 20 ml/min per 1.73 m² at 6 months and was placed on hemodialysis 11 months post-second REACT injection.

Cell Protein Marker Analysis in the REACT Bioactive Product (SRCs)

Results are summarized in Table 4. Correlation of various markers on FACS analysis revealed evidence for co-expression (Figure 2). Patients with a positive post-injection slope had a greater expression of RET with an average expression of 47.6% in the high/moderate group versus 20.6% in the low group (P = 0.027). All other cell markers were not different among patients with a positive versus a negative slope after the first injection. There was also a significant positive correlation between Six2 and RET (P = 0.0497). Correlation between various cell markers is depicted in Figure 2.

VEGF-A Analysis

VEGF-A was absent in negative control media using unconditioned serum-free renal cell growth media, but VEGF-A was detectable in SRC-conditioned, serum-free renal cell growth media (range 4.32-7.39 ng/ml) for up to 4 months.

Cell-based therapies may have the potential to modulate disease stability and offer new therapeutic options for CKD. In this cohort with moderate/severe type 2 D-CKD, percutaneously injected REACT into their kidney cortex resulted in a statistically significant improvement in eGFR decline, suggesting stabilization of renal function. eGFR, the most widely used parameter of kidney function,²⁴ improved in the moderate/high responders with a positive regression line slope. Clinical/ laboratory findings suggest that the SRCs in REACT may stabilize renal function. We are unaware of any

Table 4. Compiled data for fluorescent activated cell sorting analysis

expression is not essential for the formation of intermediate mesoderm but is essential for the differentiation toward renal and gonadal structures.²⁷ Progenitor cells are descendants of stem cells that then further differentiate to create specialized cell types. Therefore, we hypothesize that the REACT product does contain renal progenitor cells.

It is unclear whether the difference in the expression of RET is responsible for the stronger effect on eGFR. The FACS analysis suggests that membrane-bound and nuclear transcription factors representing ureteric bud, cap mesenchyme, and glomerular lineages were present in each patient's REACT product. The ureteric bud is an epithelial tube that arises from the nephric duct and branches repetitively to give rise to the kidney collecting duct system, while also generating inductive signals with the cap mesenchyme that directs mesenchymal-to-epithelial transformation and promotion of nephrogenesis by the surrounding metanephric mesenchyme cells.²⁸

SIX2 defines and regulates a multipotent self-renewing nephron progenitor population throughout mammalian kidney development, as SIX2-expressing cells give rise to all cell types of the main body of the nephron, during all stages of nephrogenesis.²⁹ During normal development, SIX2 expression and SIX2+ nephron progenitor cells in the cap mesenchyme both rapidly disappear after birth.³⁰ However, in a rat model, preserved SIX2 expression was found after partial nephrectomy in a 1-day-old neonatal rats, resulting in neonephrogenesis.³¹ The presence of SIX2 in 19 of 22 patients and a companion CM marker, OSR1, revealed lineage to the cap mesenchyme progenitor line in the SRCs.

Moreover, the LIM-class homebox transcription factor LHx1 is expressed early in the intermediate mesoderm and is one of the first genes to be expressed in the nephric mesenchyme. LHx1 is required for the specification of the renal progenitor cell field.³² Using an explant culture system to induce kidney tissue, Cirio et al.³² revealed that expression of genes from both proximal and distal kidney structures is affected by the absence of LHx1.

A key signal that promotes ureteric bud morphogenesis is GDNF, a protein secreted by metanephric mesenchyme cells that signals to ureteric bud cells by the rearrangement in transformation RET receptor tyrosine kinase, messaging the mesenchymal-to-epithelial cell transformation. In the ureteric bud and collecting ducts, RET receptor tyrosine kinase, a GDNF, and its co-receptor, GDNF family receptor α 1, initiate a signaling cascade that triggers the growth of RET-positive cells from the nephric duct toward GDNF cells of the metanephric mesenchyme.³³

Detection of RET-positive cells in our SRC population suggests the presence of a cell population capable of responding to extracellular signaling and giving rise to neo kidney-like tissue. There was significantly higher RET-positive cell expression in the patients with a positive eGFR slope after the first injection. With embryologic nephrogenesis, RET signaling, by Etv4 and Etv5, promotes competitive cell rearrangements in the nephric duct, in which the cells with the highest level of RET signaling preferentially migrate to form the first ureteric bud tip. ^33,34 RET signaling in ureteric bud cells is key for controlling cell movement, cell clustering, and ureteric bud formation during nephrogenesis. ³⁴ Whether RET expression is indeed the most important factor requires further evaluation.

Functional evaluation of SRCs provides evidence that activation of certain key developmental pathways may represent a potential mechanism of regenerative bioactivity. Molecular genetics of these developmental pathways and critical proteins that mediate nephro-genesis and their potential relevance to regeneration have been described. ¹⁷ Nephrogenesis is a dynamic cellular migration/differentiation, induced by crosstalk signaling in resident cells. It is an integral part of nephron development.

Nephrin and podocin are markers at the slit diaphragm and podocyte pedicels, respectively, and imply glomerulus lineage at the Glomerular Basement Membrane (GBM). Nephrin and the GBM form the filtration barrier , critical to repel albumin and other macromolecules from entering the Bowman's capsule, preventing epithelial inflammatory change. Glomerulogenesis is divided into the following stages: vesicle, comma- and S-shaped, glomerular capillary loop stage, and mature glomerulus,³⁵ and we have revealed these stages on histology in our animal models of SRCs.¹⁶ Animal model images closely resemble human nephrogenesis.³⁶ Furthermore, the expression of cap mesenchyme, ^37-39 ureteric,³⁴ and glomerular cell markers ^40,41 resembles in utero embryologic studies. Animal data may not completely be applicable to humans, given species evolutionary differences, but they provide a practical model. Although we have no histopathologic evidence from the patients, cell markers presented in REACT have been described as essential for nephrogenesis. The populations of SRCs contained cells responsible for nephrogenesis, and the markers analyzed represent the critical pathway for the development of the renal cortex, medullary interstitium, angioblasts, and mesangium (Table 4). These data align with our hypothesis that any improved renal function resultant from REACT injections may be due to neo kidney-like tissue development, mirroring embryonic kidney development.

In cadaveric donor-derived SRCs processed identically to REACT, we also revealed limited evidence for the expression of VEGF-A in vitro. VEGF is a proan-giogenic glycoprotein in the platelet-derived growth factor family, essential for the survival, proliferation, and differentiation of endothelial cells.^42,43 VEGF has a role in angiogenesis/vasculogenesis during embryogenesis, maintaining renal homeostasis during cell migration and is expressed throughout the life of podocytes and tubular cells.^44-46VEGF-A was expressed by donor SRCs, providing indirect evidence for angiogenesis promoting cell division, migration, endothelial cell survival, and vascular sprouting.^47,48 Our data begin to reveal a putative mechanism of action for REACT, providing the cells involved in a developmental pathway that stimulates cell migration into the diseased tissue, contributing toward neo kidney-like tissue and kidney function stabilization.

This report may serve as a proof-of-concept in humans. Additional data and analysis are required to identify which patients may benefit most from REACT therapy and to provide further evidence on the mechanism of action. Larger phase III human trials of REACT are underway (http://www.prokidney.com/clinical-trials/, accessed January 30, 2022).

Limitations

Our sample is underpowered to evaluate the impact of all covariates and bias assessment. Most patients were Caucasian and there were missing data. Our cohort is part of a larger parent study, and for this report, only 22 patients had functional SRC studies and consented for publication. Some participants had small residual cell volumes after their REACT injections, preventing full FACS analysis. There may be inherent technical limitations of the FACS analysis such as cell asynchrony. Furthermore, we only characterized the SRCs with FACS analysis after the expansion, and it is possible that this expansion process altered the results. Moreover, we did not measure CD24 and CD133 or the stem-cell specific transcription factors Oct-4 and Bml-1 and cannot confirm the presence of pluripotent stem cells responsible for formation of glomeruli as revealed by Sagrinati et al.⁴⁹

The patients with a positive slope tended to have a higher entry eGFR, and although their UACR progressed slower, the number of patients is small. We were unable to study the impact of pretreatment eGFR slope over 3 months in several patients based on the study protocol. In some patients, we had a much shorter preinjection slope (see Figure 1) which may explain why some patients had a positive slope before injection, given intrapatient variability of sCr. Nonetheless, the sustained increase in

eGFR in 7 of 22 patients is remarkable. We also do not have exogenous benchmark GFR measurements, such as iohexol. However, in our studies with 70% nephrectomized canines, SRCs increased iohexol clearance.¹⁵

Although usable SRCs could produce REACT from all biopsies, there is a possibility that the yield of usable cells from the biopsy might have been lower from patients with lower eGFR, as much of their kidney tissue could have more fibrosis, affecting FACS positivity. ^7,50 Additional studies are needed to confirm our findings and are underway.^13,14,19 Moreover, we must analyze more SRCs for the various expressions of cell line markers to determine minimum requirements for the injectable product. Further analyses with genome-wide transcriptional and epigenetic profiling of SRCs may lend support to complex developmental pathways as mechanisms of action for regenerative CKD therapies. Finally, data from animal studies cannot necessarily be extrapolated into humans and only in vitro data are presented to support the hypothesis that the GFR stabilization may be due to neo nephrogenesis.

Summary

Mechanisms of action and potency of cell therapy products are complex and recognized by regulatory agencies.⁵¹ Autologous homologous CKD cell therapies are equally complex due to multiple cell types and the many potential functional effects at the various levels in the nephron, spanning from the glomerulus to the distal convoluted tubule and collecting duct.

We have revealed that the SRCs in the REACT product evolved from preclinical trials¹² and successfully stabilized kidney function in adults with type 2 D-CKD (abstract 3611676 "Renal autologous cell therapy (REACT) for type 2 D-CKD: preliminary results with renal cortex implantation" for the American Society of Nephrology Renal Week 2021). On the basis of preclinical animal studies and this preliminary evidence, it is hypothesized that SRCs in the REACT product may function in part by promoting the assembly of progenitor cells lines, through secretion of pro-angiogenic factors, such as VEGF-A, repairing effete nephrons and possibly producing neo kidney-like tissue.

In conclusion, our trial findings suggest that the SRCs of the REACT product may initiate neo kidney-like tissue development to stabilize and improve kidney function and halttype 2 D-CKD progression. There may be benefits for a higher entry eGFR, multiple doses of REACT, and/or personalized dosing intervals to accommodate disease progression; however, the best treatment regimen remains to be elucidated.

JS, DJ, JWL, JB, RP, EB, and TB are employees of ProKidney and received wages. GF and MF received consulting fees from ProKidney. GF, JL, JB, and MF received fees for advisory roles at ProKidney.

The authors thank our REACT Trial Principal Investigators for their generous support of CKD research. ProKidney, Raleigh, North Carolina, provided funding for the trial and manuscript submission.

Data Availability Statement

Current trial efficacy data are not publicly available due to active status of the trial.

JS and TB conceived the original study idea and provided funding, intellectual insight, and supervision and editing. JS developed Tables 1 to 3. GF drafted/edited the manuscript, performed analyses, and developed the figures and tables. DJ, JL, RP, and JB developed the remaining figures. They also wrote the relevant basic science sections and conducted the in vitro experiments. MF, GF, JS and TB provided major intellectual insight into the clinical trial and the manuscript writing and editing of each version. EB and RP performed the statistical analysis. EB produced the final Figures 1 and 2. All authors contributed to and approved the final manuscript.

Supplementary File (PDF)

Appendix 1. Composition of cap mesenchyme, ureteric bud, and glomerular cell lines using cell biomarkers.

STROBE Checklist. Strobe Checklist for prospective cohort study.

Keywords

Renal autologous cell therapy · Neo-kidney augment • Chronic kidney disease • Diabetic nephropathy • Diabetes-associated chronic kidney disease • Renal outcomes

Abstract

Background: Cell therapies explore unmet clinical needs of patients with chronic kidney disease with the potential to alter the pathway toward end-stage kidney disease. We describe the design and baseline patient characteristics of a phase II multicenter clinical trial utilizing the novel renal autologous cell therapy (REACT), by direct kidney parenchymal injection via the percutaneous approach in adults with type 2 diabetic kidney disease (T2DKD), to delay or potentially avoid renal replacement therapy. Design: The study conducted a prospective, multicenter, randomized control, open-label, phase II clinical trial between an active treatment group (ATG) receiving REACT from the beginning of the trial and a contemporaneous deferred treatment group (DTG) receiving standard of care for 12 months before crossing over to receive REACT. Objectives: The objective of this study was to establish the safety and efficacy of 2 REACT injections with computed tomography guidance, into the renal cortex of patients with T2DKD administered 6 months apart, and to compare the longitudinal change in renal function between

the ATG and the DTG. Setting: This was a multicenter study conducted in major US hospitals. Patients: We enrolled eighty-three adult patients with T2DKD, who have estimated glomerular filtration rates (eGFRs) between 20 and 50 mL/min/1.73 m². Methods: All patients undergo an image-guided percutaneous kidney biopsy to obtain epithelial phenotype selective renal cells isolated from the kidney tissue that is then expanded ex vivo over 4-6 weeks, resulting in the REACT biologic product. Patients are randomized 1:1 into the ATG or the DTG. Primary efficacy endpoints for both study groups include eGFR measurements at baseline and at 3-month intervals, through 24 months after the last REACT injection. Safety analyses include biopsy-related complications, REACT injection, and cellular-related adverse events. The study utilizes Good Clinical and Manufacturing Practices and a Data and Safety Monitoring Board. The sample size confers a statistical power of 80% to detect an eGFR change in the ATG compared to the DTG at 24 months with an a = 0.05. Limitations: Blinding cannot occur due to the intent to treat procedure, biopsy in both groups, and open trial design. Conclusion: This multicenter phase II randomized clinical trial is designed to determine the efficacy and safety of REACT in improving or stabilizing renal function among patients with T2DKD stages 3a-4.

© 2022 The Author(s).

Published by S. Karger AG, Basel

Introduction

Cell therapies attempt to address unserved clinical expectations and needs of patients and medical professionals for various chronic medical conditions, including chronic kidney disease (CKD) [1, 2]. CKD affects about 18% of the world population (850 million individuals), with up to 50% secondary to type 2 diabetes which encumbers large cost expenditures, given its risk for progression to end-stage kidney disease [3-5]. Current type 2 diabetic kidney disease (T2DKD) treatments such as renal angiotensin system inhibitors and SGLT2 inhibitors utilize small molecules to modulate biochemical mechanisms of action, albeit in a damaged structural environment [6-8]. In comparison, cell therapies have the potential to induce structural nephro-restoration to stabilize or improve renal function with infrequent dosing and few or no immunologic concerns [9-11].

Most cell-based therapies under investigation for CKD utilize intravascular-delivered mesenchymal cell lines with variable dosing, potency, stability, and end-organ effect. REACT endeavors to address these shortfalls by using a personalized treatment approach with autologous, homologous cells intended to restore nephron structure, improve kidney function, and decrease CKD-relatedco-morbidities. REACT consists of selective renal cells (SRCs) isolated and expanded using Good Manufacturing Process (GMP) from kidney biopsy tissue obtained from CKD patients.

Our preclinical experience with renal autologous cells demonstrated improved renal function and nephron repair, based on surrogate markers and histologic findings in diabetic models [12]. A first-in-human phase I trial in moderate to advanced T2DKD patients confirmed cell safety and stabilization of various renal function surrogate markers during laparoscopic-guided direct renal cortex REACT delivery [13]. We herein report the design of a novel phase II open-label cell therapy trial of REACT, whereby expanded autologous renal cells are percutaneously reinjected into patients with T2DKD stages 3a-4, in a specific locoregional site of the kidney identified with computed tomography (CT) imaging. Our clinical trial design is shown in Figure 1 and is registered at the following website: https://clinicaltrials.gov/ct2/show/NCT02836574.

Our clinical trial objectives are to establish the efficacy and safety of 2 image-guided percutaneous injections of REACT into the renal cortex of patients with T2DKD, administered 6 months apart, and to compare the change in renal function between an active treatment group (ATG) and a contemporaneous deferred treatment group (DTG).

Materials and Methods

Trial Design

This is a prospective, multicenter, randomized control, open-label, phase II clinical trial. Patients were randomized in a 1:1 ratio into 2 groups (ATG n = 42 or DTG n = 41) after manufacturing confirmed the adequacy/quality of the donor kidney biopsy material and the date of REACT injection. Randomization was performed with the Interactive Web Randomization System. Since this is an open-label study, the study participants, investigators, site staff, and sponsor are not blinded to the treatment assignment. Following a kidney biopsy, patient randomization takes place into the ATG or DTG cohorts. The ATG receives REACT treatment from the beginning of the trial, whereas the DTG will receive SOC for 12 months before crossing over to receive the REACT treatment.

Participants

We included 83 patients, ages 30-80 years, who have T2DKD and an established declining estimated glomerular filtration rate (eGFR) between 20 and 50 mL/min/1.73 m2 (CKD stages 3a-4). Patients are from 17 institutions in the USA. To define the rate of CKD progression, baseline renal function was determined by 2 or more eGFR measurements, at least 3 months apart and 18 months pre-intervention. Patient demographics including clinical/laboratory characteristics and concomitant medications are shown in Table 1. The patient's blood pressure must be <150/90 mm Hg on stable doses of antihypertensives (defined as dose adjustment to no less than one-half of the current dosage or to no more than 2 times the current dosage), 6 weeks prior to the REACT injection, but dose interruptions for up to 7 days due to medical necessity are allowed. If their antihypertensives included an angiotensin-converting enzyme inhibitor or an angiotensin receptor blocker, treatment must have been initiated at least 8 weeks prior to their renal biopsy. Participants refrain from consuming nonsteroidal anti-inflammatory drugs (including aspirin), fish oil, platelet inhibitors, and anticoagulants 7 days before and after the renal biopsy and REACT injections. Inclusion and exclusion criteria are described in Table 2.

Biopsy and REACT Injection Procedures

All patients underwent an outpatient percutaneous renal biopsy and then were randomized to either the ATG or the DTG of the trial. ATG patients receive a REACT dose as soon as REACT is manufactured and delivered to the research site, whereas the DTG patients received the intervention after 12 months of SOC treatment. Patients received 2 injections of REACT given 6 months apart. In addition, each patient's rate of renal function change (based on the previous 18 months) serves as a comparator to establish the rate of DKD progression. Furthermore, the DTG patients serve only as contemporaneous comparators to the ATG patients for the 12 months during SOC treatment and against themselves once they cross over to receive REACT. The final statistical analysis will reflect the comparator time line and account for expected progression of CKD in the DTG.

A selective population of renal cells are obtained from the patient's renal cortex via percutaneous kidney biopsy, using a standard-of-clinical-care image-guided method (ultrasound or CT). Two 16-gauge or 4 18-gauge core samples >1.5 cm in length are obtained and placed immediately in the transport medium and

sent to the GMP facility for processing. An additional biopsy for local site histopathologic evaluation is allowed at the discretion of the investigator after obtaining the required cores for REACT manufacturing, depending on benefit versus risk of bleeding complications.

Selective Epithelial Renal Cell Isolation and Expansion: The cells are isolated by enzymatic digestion and expanded ex vivo, using standard cell culture techniques over approximately 4 weeks. SRCs are selected by density gradient centrifugation and formulated to produce REACT. REACT is shipped overnight as a thermolabile gelatin-based hydrogel fresh product to the clinical site, to be administered within 72 h from manufacturing release. REACT is warmed to room temperature for 30 min before the image-guided percutaneous injection into the patients' renal cortex of the same donor kidney, typically targeting a lower pole region. A

20-gauge outer guide needle (COOK Inc, Bloomington, IN, USA) is inserted into the subcapsular kidney, followed by co-axial insertion of an inner noncutting pencil tip 25-gauge needle (IMD, Huntsville, UT, USA) into the renal cortex within 5 mm of the renal capsule as shown in Figure 2. The injection target is a few millimeters deep to the subcapsular region to maximize cell deposition within the cortex microenvironment of the REACT source. The procedure is performed in an outpatient setting and with moderate sedation. Proceduralists undergo training, certification, and onsite proctoring.

The dose of REACT is 3 × 10⁶ cells/g estimated kidney weight by volumetric imaging analysis from a noncontrast MRI scan of the kidneys, obtained during patient screening. An individualized dose volume is predetermined and divided into multiple REACT deposits into the kidney. We conduct real-time assessment of nee-

Fig. 2. Percutaneous CT-guided REACT injection. Patient in prone position in CT scanner. Insertion of outer guide and inner injection needles into the renal cortex in subcapsular location (arrows). CT allows real-time evaluation of needle location, REACT deposit, and bleeding complications. REACT, renal autologous cell therapy; CT, computed tomography.

dle location and REACT deposition, and determine any perinephric bleeding during intermittent CT scanning. Post-REACT recovery care is per local site, and it includes laboratory tests and delayed renal ultrasound to assess any procedure-related complication (i.e., renal hematoma). A 24-hfollow-up clinical assessment, laboratory tests, and renal ultrasound are performed to determine renal function or delayed subclinical perinephric hematoma formation. A second REACT injection occurs approximately 6 months later into the same kidney.

Serial serum creatinine measurements are collected prerandomization and at 3-month intervals through 24 months after the last REACT injection, to calculate the eGFR by the CKD-EPI equation with serum creatinine. The rate of CKD progression for both the ATG and DTG will be compared. Additionally, each patient's disease progression rate pre-injection (eGFR slope derived from adequate historical clinical data) will be compared against the longitudinal rate of renal function decline, through the 24 months after the final REACT injection.

Randomization

We utilized a 1:1 randomization into 2 groups (ATG n = 42 or DTG n = 41) after manufacturing confirms the adequacy/quality of the donor kidney biopsy material and the date of REACT injection. The site investigator randomly assigned participants to either the ATG or the DTG using the Interactive Web Randomization System. Since this is an open-label study, the study participants, investigators, site staff, and sponsor are unblinded to the treatment assignment.

Power

Eighty-three patients who satisfied all the inclusion criteria were randomized 1:1 into either treatment group. The sample size was calculated based upon the phase I trial using an 80% power

with an a = 0.05, assuming a 33% dropout rate to detect a difference of 50% in the eGFR between groups, as the primary efficacy endpoint. The trial size also allows sufficient safety information from the procedures and REACT product to be collected.

Statistical Analysis

Utilizing eGFR slope, standard error, and sample size for each study arm, a two-sample Walsh t test will compare renal function and disease progression by CKD-EPI eGFR (2009), serum Cr, cystatin C, BUN, and urinary albumin at prerandomization, and through 24 months post-REACT injection. Patients will serve as their own control. The significance level of a = 0.05 will be set. We will evaluate the mean and standard deviation of the eGFR slope over time and at each endpoint, using a t test to determine the difference between the ATG and the DTG. Values for missing data will not be imputed. c² analysis will compare the number of patients who experience AEs in both study groups.

The complete analysis set will include all patients enrolled in the study. The injection analysis set will include all patients who receive at least 1 REACT injection. Subgroup analyses will be performed to compare the safety and efficacy data from patients who received only a single REACT injection (e.g., due to exclusion criteria occurrences, dropout, death, or other reasons) versus patients who receive 2 REACT injections. Patient disposition (including screen failure, enrolled, successful biopsy, group assignment, number of REACT injections, withdrawn pre-injection or post-injection reasons, lost to follow-up, or completed study) will be summarized for the full analysis set by frequency and the proportion of patients.

Clinical and laboratory AEs will be coded using the Medical Dictionary for Regulatory Activities System Organ Class and Preferred Term, and we will present this information summarized. Treatment-emergent AEs by seriousness, intensity, the site investigator's assessment of relationship to the kidney biopsy, REACT injection procedure or REACT product, patient discontinuation due to AEs, or deaths will be included. The number of events (occurrence) and the number of participants (incidence) who experienced AEs will be reported. The number and percent of patients who develop clinically significant laboratory abnormalities longitudinally or exhibit changes on their physical examinations will be summarized.

Clinical Outcomes

The primary objective is to assess the safety and efficacy of up to 2 percutaneous REACT injections delivered 6 months (+4 weeks) apart, into the donor kidney. The primary safety endpoint is to assess procedure- and REACT product-related adverse events through 24 months after the last REACT injection, and the primary efficacy endpoint is the measurement of eGFRs from prerandomization through 24 months after the last REACT dose. Renal-specific AEs will be monitored. Other laboratory parameters, including urinary albumin, serum BUN, and urine microalbumin/ creatinine and protein/creatine ratios, were obtained prerandomization and through 24 months after the last REACT injection. Noncontrast MRI and renal scintigraphy imaging are performed during the trial to assess morphologic change and analysis of split renal function between treated and nontreated kidneys.

Patient-Reported Outcomes

Quality of life and patient satisfaction surveys are used as subjective patient assessments. Patient-reported outcomes from the kidney disease quality of life (KDQOL) and EQ-5D-5L surveys

were obtained at baseline (defined as after randomization and before REACT injection) and through 24 months after the last REACT injection. A two-samplet test will compare mean KDQOL and EQ-5D-5L scores between groups using a significance level of a = 0.05. Each patient's baseline KDQOL and EQ-5D- 5L scores will be compared against the individual patient's scores obtained through 24 months after the last REACT injection.

Data Safety and Monitoring

An independent Data and Safety Monitoring Board (DSMB) has been chartered to oversee patient safety, especially related to unexpected investigational product-related events. The DSMB consists of 3 members with expertise directly related to protocol-specified activities. It functions independently, and its members have no other engagement with the study sponsor. The DSMB meets at regular intervals depending on the rate of patient enrollment and new data generated, and advises the sponsor on aspects concerning the safety of patients participating in the clinical trial, specifically before the patients in the DTG receive their initial REACT injection. The DSMB reviews all safety and efficacy data obtained from interim analysis and may advise on dosing plans, protocol-specified evaluations, and follow-up procedures. The DSMB shares its recommendations with the study centers, institutional review boards/ethics committees, and regulatory authorities, as appropriate. Details of specific activities and responsibilities of the DSMB are in the DSMB charter.

Discussion

T2DKD imparts a large global burden on healthcare systems and economic costs to society. Diabetes may affect 25-28% of the US population by 2050, with >40% developing CKD, resulting in high-cost comorbidities including end-stage kidney disease and renal replacement therapy [5, 14, 15]. Since the advent of renin-angiotensin system inhibitors, multiple classes of small molecule drugs have entered the clinical development pipeline to reduce DKD comorbidities, while targeting biochemical pathways that modulate metabolic, anti-inflammatory, or hemodynamic responses [6, 16, 17]. Nephroprotective effects of recent DKD therapies have demonstrated modest early reductions in declining eGFRs, although none impart improvement in the structural repair of the nephron. In addition, small molecule drugs are usually ineffective in advanced stages of DKD. Our phase II multicenter clinical trial of the novel intervention REACT will assess the safety and efficacy of autologous, homologous cell therapy in T2DKD and attempt to demonstrate nephro-restorative improvement in renal function.

Our preclinical studies with experimental models of CKD including DKD-induced and nephrectomy CKD models, showed increased survival, augmented renal function, and improved comorbidities after the cortex in-

jection of SRCs, the active biological ingredient in REACT [18, 19]. Preclinical cell markers identified an admixture of 3 cell sources (cap mesenchyme, ureteric bud, and podocyte), while lower levels of serum creatinine, blood urea nitrogen, and serum protein improved filtration correlating to a subpopulation of glomerular epithelial cells in SRCs. In addition, decreased glomerular sclerosis and mesangial proliferation in SRC-treated diabetic ZSF1 rodents were identified during necropsy at multiple interim analyses and may be predictive of positive cellular effects on multiple renal tissues in human DKD [12]. Ongoing evaluation of glomerular epithelial dysfunction in DKD may further define podocyte phenotypes that support the potential use of cellular therapies targeting renal function and proteinuria in CKD [20, 21].

Cell integration into areas of renal inflammation and fibrosis was shown by in vivo confirmation of attenuated NF-kB and PAI-1 responses to monocytic and macrophage infiltration and upregulation of tubular expansion via trophic cues [22]. Local paracrine effects by exosome secretion of cytokines modulating repair of diabetes related damage from microvesicles may also have participated in the mechanism that resulted in observed anti-fibrosis and anti-inflammatory effects, and improved renal function [12].

Cell bioactivity and biodistribution of the SRCs were confirmed by injecting the labeled active biological product into the lower or upper renal pole regions of large and small animals [12, 23]. Superparamagnetic iron oxide contrast MRI and high-content image-based immune-fluorescent analysis identified cellular dispersion throughout the kidney after locoregional injections, suggesting global implantation of cells into multiple nephrons. Gate signaling cytokines were identified in the cell mixture that elicit a chemotaxis response and mediate migration. In addition, the SRCs were identified in repaired glomeruli and tubules at necropsy utilizing immunohistology/histochemical methods and morphometric analysis [12, 23].

The SRCs of REACT biodistribution, improvement in renal function, reduced DKD comorbidities, and nonimmunogenic properties of selected renal autologous cells supported a phase I trial design. The phase I trial demonstrated feasibility and safety using hand-assisted laparoscopic access to the kidney under general anesthesia. Seven male patients underwent laparoscopic-guided cell injections with direct visualization using a large gauge needle [13]. Screening serum creatine was stable and unchanged to 18 months and increased at 24 months postinjection in 5 of 7 patients. In addition, renal clearance (calculated by Iohexol) and ACR were unchanged at 12 and

24 months, respectively, whereas cystatin C eGFR was lower at 12 and 24 months than screening levels.

Nine serious postprocedure adverse events were determined to be related to the surgical procedures with general anesthesia in the Phase I study. No conclusive adverse events were associated with the cell-based product [13].

In the present phase II clinical trial, we converted the laparoscopic injection to a CT-guided percutaneous approach, given the advances in minimally invasive procedures with smaller-platform medical devices, imaging methods, and conscious sedation in an outpatient setting. These changes will mitigate adverse procedural events, and image guidance will ensure REACT enters the renal cortex. This approach currently is also used in a phase I open label trial of CKD due to congenital anomalies of the urinary tract [24].

Conclusions

REACT holds the promise to achieve improved renal function by nephron-repair and restoration of the diseased kidney with its unique cell composition and delivery method. The anti-fibrotic, anti-inflammatory, and non-immunologic features of REACT may ameliorate the progression of DKD and reduce the need for renal replacement therapy. An improved structural function could complement and enrich the bioactivity of other small molecules affecting T2DKD renal function and comorbidities.

Acknowledgements

We wish to thank our REACT trial participants for their generous support of CKD research. We also would like to thank Ms. Brenda McGrath for statistical management and analysis, Ms. Chelsey Hehl for data administration assistance, and Ms. Beth Hilburger for manuscript preparation, all from ProKidney.

Statement of Ethics

The trial protocol was approved by each site's institutional review board or ethics committee on human research, and the participants provided written informed consent. The trial identifier number is NCT:02836574 and found at https://clinicaltrials.gov/ ct2/show/NCT02836574.

Conflict of Interest Statement

J. Stavas, A. Johns, D. Jain, and T. Bertram are employed by ProKidney. S.G. Coca has received fees for advisory boards or steering committee roles for Renalytix, CHF Solutions, Bayer, Boehringer Ingelheim, Takeda, Vifor, Quark, ProKidney, and Akebia in the past 3 years. He owns equity in Renalytix, and receives salary and research support from Renalytix, ProKidney, XORTX, and the Renal Research Institute. A. Silva and M. Díaz-González de Ferris received fees for advisory board roles for ProKidney.

Funding Source

ProKidney, Grand Cayman, Cayman Islands, provided funding for the trial and manuscript submission.

Author Contributions

J.S. and M.F. made substantial contributions to the design of the work, drafting, revising, and final approval of the version to be published and accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. A.S., D.G., S.C., A.J., D.J., and T.B. contributed to revisions and final approval of the version to be published. G.B. contributed to trial design and final approval.

Data Availability Statement

Current trial safety and efficacy data are not publicly available due to active status of the trial.