Rationale and Design of the CONCERT-HF (Combination Of meseNchymal and c-kit⁺ Cardiac stEm cells as Regenerative Therapy for Heart Failure) Trial

Abstract

Rationale

Autologous bone marrow (BM) mesenchymal stem cells (MSCs) and c-kit⁺ cardiac progenitor cells (CPCs) are two promising cell types being evaluated for patients with heart failure (HF) secondary to ischemic cardiomyopathy. No information is available in humans regarding the relative efficacy of MSCs and CPCs and whether their combination is more efficacious than either cell type alone.

Objective

CONCERT-HF (Combination Of meseNchymal and c-kit⁺ Cardiac stEm cells as Regenerative Therapy for Heart Failure) is a Phase II trial aimed at elucidating these issues by assessing the feasibility, safety, and efficacy of transendocardial administration of autologous MSCs and CPCs, alone and in combination, in patients with HF caused by chronic ischemic cardiomyopathy (coronary artery disease and old myocardial infarction).

Methods and Results

Using a randomized, double-blinded, placebo-controlled, multi-center, multi-treatment, and adaptive design, CONCERT-HF examines whether administration of MSCs alone, CPCs alone, or MSCs + CPCs in this population alleviates left ventricular (LV) remodeling and dysfunction, reduces scar size, improves quality of life, or augments functional capacity. The four-arm design enables comparisons of MSCs alone with CPCs alone and with their combination. CONCERT-HF consists of 162 patients, 18 in a safety lead-in phase (Stage 1) and 144 in the main trial (Stage 2). Stage 1 is complete and Stage 2 is currently randomizing patients from seven centers across the US.

Conclusions

CONCERT-HF will provide important insights into the potential therapeutic utility of MSCs and CPCs, given alone and in combination, for patients with HF secondary to ischemic cardiomyopathy.

INTRODUCTION

Cell therapy has recently emerged as a potential novel approach to the treatment of heart failure (HF) caused by chronic ischemic cardiomyopathy (coronary artery disease and old myocardial infarction), hereby referred to as “ischemic HF”. Many studies have explored the use of various types of stem or progenitor cells in patients with chronic ischemic HF¹, with encouraging results^1–4. However, since the studies performed to date have been relatively small, definitive demonstration of therapeutic efficacy awaits the results of large, well-designed Phase III trials.

Two of the most promising types of cells being considered for patients with ischemic HF are bone marrow (BM) mesenchymal stem cells (MSCs)⁴ and c-kit⁺ cardiac progenitor cells (CPCs)^{1, 5}. No information is available regarding the relative efficacy of MSCs and CPCs in humans. MSCs have been shown to ameliorate post-myocardial infarction (MI) left ventricular (LV) dysfunction and reduce scar size in preclinical models of ischemic cardiomyopathy (ICM), including rodents⁶ and pigs^7–14. Several clinical trials also have suggested that MSCs are safe and therapeutically beneficial in patients with chronic ischemic HF^{4, 15–23}. Similarly, CPCs have been found to attenuate LV dysfunction and remodeling in many preclinical studies from several independent laboratories, both in rodent and in porcine models of chronic ICM^24–29. A Phase I trial in patients with ischemic HF has provided evidence for the safety and feasibility of administering autologous CPCs after bypass surgery and has suggested benefit, although the study was not designed to establish therapeutic efficacy^{30, 31}.

Another important question is whether the combination of MSCs and CPCs is more efficacious than either cell type alone. A beneficial interaction between these cells types is suggested by preclinical evidence that transplantation of MSCs promotes proliferation of endogenous CPCs, suggesting that this may be an important mechanism for the salutary effects of MSC administration⁷. Indeed, when MSCs and CPCs were combined in preclinical models, the therapeutic effects were additive, and thus superior to those of either cell type alone^{24, 25}. The concept of combinatorial stem cell therapy is conceptually attractive but has never been tested in humans.

CONCERT-HF (Combination Of meseNchymal and c-kit⁺ Cardiac stEm cells as Regenerative Therapy for Heart Failure) (clinicaltrials.gov Identifier: NCT02501811) is a Phase II trial aimed at addressing the aforementioned issues by assessing the feasibility, safety, and efficacy of MSCs and CPCs, alone and in combination, in patients with ischemic HF. Using a randomized, double-blinded, placebo-controlled, multi-center, adaptive design, CONCERT-HF addresses several important questions: Is combined treatment with autologous MSCs and CPCs feasible and safe in patients with ischemic HF? Do MSCs and CPCs, given alone or in combination, alleviate LV dysfunction, reduce scar size, improve quality of life, and/or augment functional capacity in this population? Is either cell type more effective than the other? Is the combination of MSCs and CPCs superior to MSCs alone or CPCs alone in terms of therapeutic efficacy?

METHODS

Study population

Upon completion of the CONCERT-HF trial, results will be publicly available at www.clinicaltrials.gov and data will be available by request to the corresponding author of the final results paper. Inclusion and exclusion criteria are summarized in Tables 1 and 2, respectively. Candidates for enrollment include men and women, aged 21 to 79, with HF of ischemic etiology. Participants must have documented coronary artery disease, evidence of myocardial scar involving ≥5% of the LV mass by magnetic resonance imaging (MRI), LV systolic dysfunction (LV ejection fraction [LVEF] ≤40% by MRI), and symptoms of HF (NYHA class II-III), and must be receiving guideline-driven medical therapy (e.g., beta blockers, ACE inhibitors/ARBs/ARNIs, and/or aldosterone antagonists) at stable and tolerated doses for ≥1 month before consent. At least 3 months must have elapsed from a percutaneous coronary intervention or implantation of a CTR device.

Table 1.

Inclusion Criteria

To participate, a patient MUST: Be ≥ 21 and <80 years of age Have documented coronary artery disease with evidence of myocardial injury, LV dysfunction, and clinical evidence of HF Have a “detectable” area of myocardial injury defined as ≥ 5% LV involvement (infarct volume) and any subendocardial involvement by cMRI Have an EF ≤ 40% by cMRI Be receiving guideline-driven medical therapy for heart failure at stable and tolerated doses for ≥ 1 month prior to consent Be a candidate for cardiac catheterization Have NYHA class I, II or III heart failure symptoms If a female of childbearing potential, be willing to use one form of birth control for the duration of the study, and undergo a pregnancy test at baseline and within 36 hours prior to injection

Table 2.

Exclusion Criteria

To participate, a patient MUST NOT HAVE: Indication for standard-of-care surgery (including valve surgery, placement of LV assist device, or imminent heart transplantation), coronary artery bypass grafting (CABG) procedure, and/or percutaneous coronary intervention (PCI). Candidates cannot be UNOS 1A or 1B, and they must have documented low probability of being transplanted. PCI within 3 months of randomization CABG within 4 months of randomization Valvular heart disease including mechanical or bioprosthetic heart valve, severe valvular (any valve) insufficiency/regurgitation within 12 months of consent, and aortic stenosis with valve area ≤ 1.5 cm2 History of ischemic or hemorrhagic stroke within 90 days of consent History of an LV remodeling surgical procedure utilizing prosthetic material Presence of a pacemaker and/or ICD generator with any of the following limitations/conditions: ○ manufactured before the year 2000 ○ leads implanted < 6 weeks prior to consent ○ non-transvenous epicardial or abandoned leads ○ subcutaneous ICDs ○ leadless pacemakers ○ any other condition that, in the judgment of device-trained staff, would deem an MRI contraindicated Pacemaker-dependence with an ICD (pacemaker-dependent candidates without an ICD are not excluded) A cardiac resynchronization therapy (CRT) device implanted within 3 months of consent Other MRI contraindications (e.g., patient body habitus incompatible with MRI) An appropriate ICD firing or anti-tachycardia pacing (ATP) for ventricular fibrillation or ventricular tachycardia within 30 days of consent Ventricular tachycardia (≥ 20 consecutive beats) without an ICD within 3 months of consent, or symptomatic Mobitz II or higher degree atrioventricular block without a functioning pacemaker within 3 months of consent Presence of LV thrombus Baseline VO2 max > 75% of age and gender based predictive values Baseline eGFR < 35 ml/min/1.73 m2 Poorly controlled blood glucose levels (HbA1c > 10%) Hematologic abnormality evidenced by hematocrit < 25%, white blood cell < 2,500/ul or platelet count < 100,000/ul Liver dysfunction evidenced by enzymes (AST and ALT) ˃ 3 times the ULN Coagulopathy (INR ≥ 1.3) not due to a reversible cause (e.g., warfarin and/or Factor Xa inhibitors). Patients who cannot be withdrawn from anticoagulation will be excluded. HIV and/or active HBV or HCV Allergy to radiographic contrast material that cannot adequately be managed by premedication Known history of anaphylactic reaction to penicillin or streptomycin Received gene or cell-based therapy from any source within the previous 12 months History of malignancy within 5 years, excluding basal cell carcinoma and cervical carcinoma in situ which have been definitively treated Condition that limits lifespan to < 1 year History of drug or alcohol abuse Chronic immunosuppressant therapy such as corticosteroids or TNFα antagonists Cognitive or language barriers that prohibit obtaining informed consent or any study elements Pregnancy or lactation or plans to become pregnant in the next 12 months Any other condition that, in the judgment of the Investigator or Sponsor, would impair enrollment, study product administration, or follow-up

Study design

The primary objectives are to assess: i) whether autologous MSCs and CPCs, alone or in combination, can be manufactured and delivered to patients with ischemic HF (feasibility); ii) whether the administration of MSCs and CPCs, alone or in combination, is well-tolerated (safety); and iii) whether MSCs and CPCs, delivered alone or in combination, improve LV function, quality of life, and functional capacity, and/or reduce scar size (efficacy).

The planned total sample size is 160 patients, enrolled in two stages. In Stage 1 (open label, lead-in study) 16 patients (expanded to 18) were randomized 1:1 to either a standard-of-care control group (i.e., they did not undergo harvest, mapping, or injection procedures) or combination therapy (MSCs + CPCs, as described below for Stage 2). Stage 1 participants were followed for three months to complete safety and functional assessments. After review of the three-month data by an NHLBI Data and Safety Monitoring Board (DSMB), approval was granted to proceed with Stage 2 (vide infra). Patients randomized to MSCs + CPCs in Stage 1 were followed for 12 months for safety. Those randomized to the standard-of-care control group had the option to be evaluated for enrollment in Stage 2.

Stage 2 is a Phase II, double-blind, randomized, placebo-controlled trial designed to evaluate the feasibility, safety, and efficacy of MSCs alone, CPCs alone, and their combination compared with placebo as well as each other in patients with ischemic HF (Figure 1). Using an adaptive design, Stage 2 patients (n=144) are randomized (1:1:1:1) to one of four treatments: placebo (cell-free PlasmaLyte-A), autologous MSCs (“MSCs alone”; target dose, 150 × 10⁶ cells), autologous CPCs (“CPCs alone”; target dose, 5 × 10⁶ cells), or a combination of autologous MSCs and CPCs (“MSCs + CPCs;” target doses, 150 × 10⁶ MSCs plus 5 × 10⁶ CSCs).

Figure 1.

Protocol Overview for Stage 2.

Baseline testing, randomization, and tissue harvesting

Within 60 days of written informed consent, patients undergo baseline testing (Table 3). Any eligible patients are randomized via an online database created and maintained by the Data Coordinating Center (DCC) and undergo, on the same day, bone marrow aspiration (BMA) and right heart catheterization (RHC) (all groups) and right ventricular endocardial biopsy (EMB) (only in the CPCs alone and MSCs + CPCs groups).

Table 3.

Baseline Evaluation

Comprehensive review of medical and surgical history, vital signs, and physical examination Current prescription and over-the-counter medications New York Heart Association Classification and Canadian Cardiovascular Society Functional Classification Laboratory tests, including: chemistry tests, complete blood count, liver function tests, NT-proBNP, troponin I/T, HbA1c, PT/INR/PTT, and pregnancy test (childbearing women) Infectious disease tests (institution bone marrow donor panel) including: HIV, Hep B (HBsAg, Anti-HBs, Anti-HBc), and Hep C (Anti-HCV) Cardiac magnetic resonance imaging and implantable cardioverter defibrillator (ICD) interrogation (if applicable) Six-minute walk test 12-lead electrocardiogram Treadmill-based VO2 max Minnesota Living with Heart Failure Questionnaire Sexual function assessment: International Index of Erectile Function for males; Female Sexual Function Index for female s

NT-Pro BNP= N-Terminal pro-Brain Natriuretic Peptide; HbA1c= Glycated Hemoglobin (A1c); PT/INR/PTT= Prothrombin Time/International Normalized Ratio/ Partial Thromboplastin Time; HIV=Human Immunodeficiency Virus; HBsAg=Hepatitis B surface antigen; Anti-HBs=Hepatitis B surface antibody; Anti-HBc=Hepatitis B core antibody; Anti-HCV= Hepatitis C Virus

A total of ~90 ml of BM is harvested from the posterior iliac crest. Upon arrival at the local cell processing laboratory, BM samples undergo quality control (QC) tests including sterility (aerobic, anaerobic, and fungal cultures), nucleated cell count, and viability (Trypan Blue). Fresh BM samples are shipped to both the Central Cell Manufacturing Facility (CCMF) for production of MSCs and, with appropriate consent, to the CCTRN Biorepository (BRC)³².

In patients assigned to the CPCs alone and MSCs + CPCs groups, EMB is performed during RHC. Up to six endomyocardial samples are obtained to achieve approximately 20 mg total tissue. Fresh EMB samples are shipped to the CCMF for isolation and expansion of CPCs. Two transthoracic 2-D echoes without contrast are done on the day of harvest for all study patients: 1) a pre-RHC procedure echo to assess for pre-existing pericardial effusion and 2) a post-RHC procedure performed within 6 hours post-procedure. To minimize the risk of serious adverse events, RHC procedures are canceled or terminated if the patient’s NYHA class deteriorates to class IV, if the right ventricular or pulmonary artery systolic pressure is ≥60 mmHg, or if the pulmonary capillary wedge pressure is ≥35 mmHg. Furthermore, the RHC procedure is temporarily halted and the suitability of the patient for continuation re-evaluated if systemic blood pressure is <80 mmHg (change from baseline), heart rate is >100 beats/min (change from baseline), either right ventricular or pulmonary artery systolic pressures are 50-59 mmHg, or pulmonary capillary wedge pressure is 30-34 mmHg.

Cell manufacturing

MSCs and CPCs are manufactured by the CCMF at the University of Miami Miller School of Medicine’s Interdisciplinary Stem Cell Institute.

MSCs are manufactured using each patient’s own BM. The mononuclear cells (MNCs) are isolated using a density gradient with Lymphocyte Separation Medium. The low-density cells are collected from the gradient and plated for further expansion. After 14 days of culture, passage zero (P0) cells are harvested by enzymatic treatment and expanded into 60 flasks. These flasks are incubated for a week and the MSCs are harvested again (P1 cells). The P1 cells are then washed and total viable cell counts are determined for cell dosing. Samples of MSC products are characterized to assure that they meet predetermined specifications with regard to cell count, phenotype, potency, mycoplasma testing, and other QC tests (Table 4). MNCs of participants randomized to CPCs alone or to placebo are cryopreserved and shipped to the BRC. Media from the final MSC passage prior to product cryopreservation is likewise cryopreserved and sent to BRC.

Table 4.

Quality Control Testing of Cell Product Prior to Cryopreservation at the CCMF

Assay	Test Method	Specification
Mycoplasma PCR	VenorGeM® (Cells in Conditioned Medium prior to cryopreservation)	Negative
Viability	Trypan Blue	≥70%
Aerobic, Anaerobic, and Fungal	14 day Bactec/BacT/ALERT assay or equivalent	No Growth
MSC Product only:
CFU-F	Colony Formation, 14 Days	Growth
Phenotype/Cell Characterization CD 105+ CD45+	Flow Cytometry	CD105+ >80% CD45+ <2%
Cell Count	Hemocytometer	>75×106
CPC Product only:
Phenotype/Cell Characterization CD 117+ CD45+	Flow Cytometry	CD117+ ≥70% CD45+<2%
Cell Count	Manual	>0.8 ×106

PCR: polymerase chain reaction; CFU-F: colony forming unit fibroblasts.

In patients randomized to the CPCs alone or the MSCs + CPCs groups, CPCs are manufactured using each patient’s own EMB samples. The methods for culture and subsequent production of CPCs are a modified version of the protocol previously used for SCIPIO³⁰. Briefly, EMBs are enzymatically digested with warm Collagenase Type II (Gibco # 17101) solution (final concentration 329 U/ml diluted in Ham’s F12) at 37°C. The digested solution is collected, washed, and plated in culture media to expand CPCs. Cells are expanded in culture for approximately 3-4 weeks; when cell confluence >80% is attained, the culture is enriched for c-kit⁺ cells using immunomagnetic beads (CD117 MicroBead Kit, Miltenyi Biotec). The c-kit⁺ CPCs are further expanded to generate the CPC product. Samples of the cell suspension taken from the last passage are characterized to assure that they meet predetermined specifications, including cell count, phenotype (CD117 positivity), mycoplasma testing, and other QC tests, as indicated in Table 4. Isolated CPC products are then suspended in cryoprotectant at 0.5-1.0×10⁶ cells/ml and cryopreserved using a control rate freezer. The frozen vials are placed into the vapor phase of liquid nitrogen freezers and stored for at least three weeks. These products are not released until the required sterility testing has been completed on samples obtained before and after adding cryopreservation medium as outlined in Tables 4 and 5. Media from the final CPC passage prior to product cryopreservation is likewise cryopreserved and sent to BRC.

Table 5.

Quality Control Tests on Thawed, Washed Cell Product at Center Cell Processing Labs

Assay	Test Method	Specification
Rapid Sterility*	Gram Stain	No organisms seen (negative)
Viability*	Trypan Blue	≥70%
Endotoxin*	EndoSafe PTS	≤ 5EU/kg†
Aerobic, Anaerobic, and Fungal	14 day Bactec/BacT/ALERT assay or equivalent	No Growth
Cell Count – CPCs*	Manual	Minimum 0.8×106 final dose Maximum 5×106 final dose
Cell Count – MSCs*	Manual	Minimum 75×106 final dose Maximum 150×106 final dose

^*Tests done for release criteria specification prior to injection; for cell counts, the product must meet the minimum count for release

^†Based on recipient weight and product volume

MSC and CPC products are cryopreserved at the CCMF and stored in liquid nitrogen until ready to be shipped to the local cell processing labs. Cell products are transported to each clinical site via validated liquid nitrogen dry shipper within one week prior to the scheduled injection procedure. Local cell processors receive and prepare all products for administration. Placebo product consists of cell-free PlasmaLyte-A. All study products (MSCs alone, CPCs alone, MSCs + CPCs, and placebo) are provided to the treatment teams in a packaging with identical appearance.

Biorepository core (BRC)

An optional central BRC is utilized for participants who provide additional informed consent. The goal is three-fold: i) to provide storage of critical biomaterials (i.e., participant peripheral blood, BM, MSC product, and CPC product), ii) to provide long-term integrity (up to 10 years) of these biospecimens and products, and iii) to provide management of immunologic, immunohistochemical, cellular, and molecular analyses of collected samples, as well as phenotypic and functional analyses of cells and plasma samples. The aim of the BRC is to identify factors that modulate disease progression and/or predict successful intervention and to gain insights into potential mechanisms of action of MSCs and CPCs as well as the relationship of MSC and CPC phenotypes to clinical outcomes. Samples collected and available for further analyses via ancillary studies are specified in Table 6.

Table 6.

Biorepository

Biologic	Sample #	Sample details	Potential ancillary analyses on stored (frozen) samples†
Bone marrow	n=144 (all patients)	Fresh BM, flash frozen buffy coat, cryopreserved buffy coat, cryopreserved BM supernatant	In vivo and in vitro studies, RNA, DNA, protein
BM-MNCs	n=72, (placebo & CPC patients)	Cryopreserved MNCs	In vivo and in vitro studies, RNA, DNA, protein
Excess cell product	n=144 (3 active groups)*	10-30 million MSCs, 0.5-1 million CPCs	In vivo and in vitro studies, RNA, DNA, protein
Cell culture medium	n=144, (3 active groups)*	up to 10 12-ml aliquots	Secretome (microparticles/ exosomes, miRNAs, cyto/chemokines, GFs)
Peripheral Blood	n=144 (all patients)	Flash frozen buffy coat and plasma, cryopreserved buffy coat and plasma	In vivo and in vitro studies, RNA, DNA, protein, secretome (microparticles/ exosomes, miRNAs, cyto/chemokines, GFs)
Explanted Heart (as available)	Unknown	10 cardiac regions, formalin fixed or flash frozen	RNA, DNA, protein, histology,

^*One product/MSC patient, 1 product/CPC patient; 2 products/MSC + CPC patient

^†Analyses require ancillary study application and funding source outside scope; samples are stored in such a way as to realize these potential future analyses

Study product delivery

Approximately 14 weeks after harvest, participants return to the cardiac catheterization laboratory for cardiac mapping and to receive the study product. The product undergoes cell counts and QC testing before injection, as indicated in Table 5. Each patient receives 15 intramyocardial, electromechanically-guided injections using the NOGA® XP Mapping and Navigation System (Biologic Delivery Systems, McKees Rocks, PA). Intervention groups receive one of three treatments: MSCs in 6 ml of PlasmaLyte-A (target dose, 150×10⁶ cells; minimum dose, 75×10⁶ cells), CPCs in 6 ml of PlasmaLyte-A (target dose, 5×10⁶ cells; minimum dose, 0.8×10⁶), or MSCs + CPCs (same doses as above) in 6 ml of PlasmaLyte-A. The placebo group receives 6 ml of cell-free PlasmaLyte-A. In order to maintain study blinding, if a patient randomized to MSCs + CPCs does not meet the minimum dose of CPCs, he/she will receive MSCs alone and conversely, if a patient does not meet the minimum dose of MSCs, he/she will receive CPCs alone; should both products fail to meet release criteria, the patient would receive placebo. If a patient in either the MSCs or CPCs alone groups does not meet the minimum dose, he/she will receive placebo. For statistical analysis purposes, all participants will be analyzed in the group to which they were randomized in accordance with the “intention-to-treat” principle.

Injections are targeted to the border zones (5-10 mm regions adjacent to scar) identified by electromechanical mapping. Injections are placed so as to encircle the scar, with injection sites in both i) the viable border zone surrounding the area of scar and ii) the scar adjacent to the viable border zone. Priority is given to viable zones with a unipolar voltage ≥7 mV and a premature ventricular contraction on extension of the needle; some injections are made in the border zone with a voltage ≥4mV without the requirement of a premature ventricular contraction. In addition, injection sites should satisfy the following two criteria: i) perpendicular position of the catheter to the LV wall, and ii) loop stability <4 mm. After proper placement of the catheter is confirmed, patients in the treatment groups receive a needle injection of 0.4 ml of solution (5-10 × 10⁶ MSCs and/or 0.05-0.33 × 10⁶ CPCs per injection) delivered over 60 s; those in the placebo group receive vehicle (0.4 ml of PlasmaLyte-A per injection). All patients are monitored overnight. A transthoracic 2-D echo without contrast is performed within 6 h after the procedure to evaluate for pericardial effusion, and blood samples for troponin measurements are collected ~ 8 h following the procedure and prior to discharge.

Outcomes

Visits occur at 1 day, 1 week, and 1, 3, 6, and 12 months after study product injection. A telephone contact takes place at 24 months to assess the patient’s current medications, as well as morbidity and mortality. The primary endpoints assess feasibility, safety, and efficacy (Table 7). All adverse events grade 2 and higher, based on the Common Terminology Criteria for Adverse Events, are captured. These events include major adverse cardiac events related to HF (HF-MACE) (death, hospitalization for worsening HF, and/or exacerbation of HF not requiring hospitalization) and other significant clinical events. The feasibility of harvests, preparing and delivering the intended number of cells, and collecting cardiac MRI variables in patients with cardiac devices, is assessed.

Table 7.

Study Endpoints

Feasibility assessment (the following measures will be reported)
Number and percent of patients who have: Events between randomization and study product injection (SPI) that preclude the procedure Failed bone marrow aspiration procedure Failed endomyocardial biopsy procedure Failed release criteria (including minimum number of cells) for receiving the MSC product Failed release criteria (including minimum number of cells) for receiving the CPC product Less than 15 injections during the SPI procedure At least one cardiac MRI endpoint measure that is uninterpretable due to issues related to the device, including, but not limited to, inability to undergo the procedure
Safety assessment (assessed from baseline to 6 months and from baseline to 12 months)
Major Adverse Cardiac Events (HF MACE) Death Hospitalization for worsening heart failure Exacerbation of heart failure (not requiring hospitalization)
Other Significant Clinical Events Non-fatal stroke Non-fatal myocardial infarction Coronary artery revascularization Ventricular tachycardia/fibrillation Pericardial tamponade
All adverse events at least grade 2 in severity (Common Terminology Criteria for Adverse Events)
Prospectively declared efficacy endpoint measures: (assessed from baseline to 6 months and from baseline to 12 months)
Myocardial Function and Structure (cardiac MRI) Function: Change in LV ejection fraction, global strain, and regional strain Structure: Change in LV end-diastolic and end-systolic volume indices and LV sphericity index Morphology: Change in infarct/scar volume
Functional Capacity Change in VO2 max (treadmill) Change in exercise tolerance (6-minute walk test) Change in MLHF Questionnaire score
Clinical Outcomes MACE Cumulative days alive and out of hospital
Biomarkers Change in NT-proBNP

Efficacy measures are obtained at baseline and at 6 and 12 months after study product injection, followed by a 24-month phone call. To comprehensively assess the efficacy of cell therapy, multiple endpoints have been selected from different categories of effects (domains) (Table 7). MRI evaluations of LV function, LV structure (including LV volumes), and scar burden provide a comprehensive assessment of myocardial performance and morphology. MRI evaluations are performed at the time of screening, within 30 days prior to study product injection (baseline), and at 6 and 12 months after injection. The presence of a pacemaker or implanted defibrillator device is not a contraindication to MRI^{33, 34}; in these patients, the procedures recommended by the American Heart Association³³ are followed. All MRI analyses are performed in a core laboratory by investigators masked to treatment assignment and other patient data. Functional capacity encompasses maximal oxygen consumption (VO₂ max, treadmill test) (also assessed by a core laboratory blinded to treatment assignment), 6-min walk test, and quality of life assessment (Minnesota Living with Heart Failure Questionnaire [MLHFQ]). We also assess clinical outcomes (measured by HF-MACE and cumulative days alive and out of the hospital) and biomarkers of heart failure (NT-proBNP).

Statistical methods

The principal analysis will be based on an “intention-to-treat” assessment of the three cell therapies. The reasons for non-treatment of the anticipated small number of participants who are randomized but do not receive any treatment, or do not receive the randomly assigned treatment, will be tabulated. An “as-treated” evaluation placing patients in treatment groups based on what they received also will be conducted. Safety and feasibility data will be reported for participants in both the Stage 1 lead-in study (n=18) and the Stage 2 main cohort study (n=144). Exact testing for categorical variables and analysis of variance for continuous variables will be used to evaluate the differences in baseline variables between treatment groups.

The change in every endpoint variable will be assessed using a repeated-measures general linear model for i) the 6-month follow-up and ii) the 12-month follow-up. The dependent variable in this model will be the change in the treatment group; the independent variable will be the therapy group assignment. In addition, an interaction term consisting of the product of the CPCs and MSCs alone treatment arms will be included to assess the presence of cell therapy synergism. These analyses will provide the effect of each treatment for each assessment time period for each prospectively declared endpoint. Because the effects of MSCs and CPCs in these patients are unknown, and because this is not a pivotal trial, no adjustment for multiplicity will be used. This comprehensive analysis is based on the concept that early phase studies should provide the most complete and elaborative exposition and synthesis of data in order to guide the direction of future cell therapy research³⁵. Sample sizes were chosen assuming a 20% attrition rate. Safety and feasibility data from Stage 1 will be included in the formal analysis of the study.

Underestimation of the sample sizes needed to achieve the desired power would weaken the conclusions of the study, whereas overestimation would cause unnecessary expense of time and resources. Consequently, CONCERT-HF uses an adaptive design for finalizing the sample size. Specifically, the final sample size will be decided on the basis of the standard deviation (SD) observed for each of the change in EF and of the change in infarct size between the baseline and 6-month data points in the control group. Two interim analyses will be performed when 34% and 66% of the control group have completed the 6-month data collection and the data will be submitted to the DSMB for review and recommendation.

RESULTS

Study design

CONCERT-HF is a Phase II trial that assesses the effects of MSCs and CPCs, given alone or in combination, in patients with HF of ischemic etiology. The four-arm design enables this trial to address multiple questions: i) Does administration of MSCs alone have beneficial effects on LV dysfunction and remodeling, scar size, functional capacity, quality of life, and/or clinical outcome? ii) Does administration of CPCs alone have beneficial effects on these variables? iii) Does administration of MSCs + CPCs have beneficial effects on these variables? iv) Is the combination of MSCs and CPCs superior to either cell type alone? v) When compared to each other, is either cell type alone superior? Each of these questions has obvious clinical significance. Considerable new information will be provided by CONCERT-HF. The effects of CPCs in ischemic HF have been evaluated in a Phase I, single-center trial^{30, 31} but not in multicenter, randomized, double-blind studies powered for efficacy. The relative effects of MSCs and CPCs in patients have not been compared in the same trial. Further, no previous study has evaluated a combination of two different cell types.

Changes in study design resulting from Stage 1 (lead-in study)

Stage 1 was conducted to assess the feasibility and safety of the study procedures as well as the bioactivity of the products. One patient enrolled in Stage 1 died during the EMB procedure because of right ventricular perforation resulting in hemopericardium and tamponade. In a study of patients with severe HF who undergo invasive procedures, adverse events are expected³⁶. However, as a result of the unfortunate death during Stage 1, several important changes to the design were made prior to the initiation of Stage 2 to further reduce risk to study participants. The study protocol was modified as follows: i) the EMB procedure was replaced with a sham procedure (RHC without EMB) in patients allocated to the control and MSCs alone groups; ii) stopping and/or halting thresholds were added on the day of harvest based on RHC pressures, heart rate, and NHYA class; iii) more safety guidance was added for EMB harvesters, and harvesters are required to participate in ongoing best practices sessions aimed at enhancing the safety of the procedure; iv) pre- and post-EMB echocardiograms were added to detect new pericardial effusion; and v) a post-consent case review by the Steering Committee was implemented for each potential participant prior to final eligibility determination. In addition, several procedural changes were made to maximize the ability to consistently reach the target CPC dose of 5×10⁶ cells.

Blinding of investigators

As noted above, Stage 2 includes sham EMB procedures for half of the study population (patients assigned to placebo or MSCs alone groups). In our judgement it is unlikely that the EMB has any impact on patient outcomes, particularly because the study product is administered ~ 3.5 months later. All patients undergo BMA, RHC, and NOGA-guided injections. A sham protocol is followed for patients undergoing RHC without EMB, so that they are not aware of whether an EMB is performed or not. Double-blinding is maintained by having separate blinded and unblinded study teams at each center; the team members directly involved in RHC/EMB procedures are not involved in any other part of the protocol, and all other members of the team are blinded to the therapy assignment. Similarly, cardiac MRI and VO₂ max endpoints are determined by core laboratories that are blinded to therapy assignment.

Management of variability in MRI endpoints

The power of CONCERT-HF is very sensitive to the variability of its endpoints. For example, the power to detect an increase of 6 absolute EF units is 81% if the standard deviation (SD) is 8, but drops to 72% if the SD increases by only 1 unit to 9. To minimize variability of MRI-based endpoints, the DCC and the MRI Core Lab monitor the SDs of individual centers (using only data in the control group). If center outliers are identified with respect to the SD of MRI endpoints, an effort will be made to reduce such variability, e.g., by assessing technicians’ ability to obtain images and providing additional training when needed, by evaluating changes in equipment at each center and/or need for software/hardware updates, by monitoring the rate of staff turnover and its impact, etc.

Group sizes and adaptive design

One of the most innovative features of CONCERT-HF is the use of an adaptive design with regard to sample sizes (see Statistical methods). The sample sizes described herein are tentative, and were selected to ensure sufficient power to detect meaningful changes in endpoint measures. Power calculations were based on the variability encountered in previous trials in similar patient populations, including SCIPIO^{30, 31} and TIME³⁷, LateTIME³⁸, and SWISS-AMI³⁹. According to these estimates, CONCERT-HF is adequately powered to detect significant changes in the following endpoints: a five unit increase in LVEF (power, 89%), a 4.5% decrease in myocardial scar size (93%), a 12 ml decrease in LV end systolic volume (LVESV) (86%), and a 0.08 change in sphericity index (86%). Power for our other prospectively declared endpoints (i.e., change in LV end diastolic volume [LVEDV], VO₂ max, 6-minute walk test, and MLHFQ score) is marginal at 70-80%, but given the anticipated interest in these outcomes, we are committed to reporting findings on these variables to the research community. In accordance with the adaptive design of the trial, final group sizes will be determined on the basis of two interim analyses (see Statistical methods).

DISCUSSION

Study design

Because data on the safety and feasibility of MSCs and CPCs have been reported previously^{16, 18, 30, 31}, CONCERT-HF was designed as a Phase II trial to assess not only feasibility and safety but also efficacy, and to allow comparisons of each of the four treatment arms to the other three. Selecting an appropriate primary endpoint, however, is difficult because the effects of MSCs and CPCs and their mechanism of action in humans remain unclear. Although LVEF has been used as the primary endpoint in many trials, failure to improve LVEF may not necessarily signify absence of benefit, as exemplified by recent studies in which cell therapy reduced MACE^{21, 23} or improved the 6-min walking distance and MLHFQ score^{16, 18} without improving LVEF. Accordingly, CONCERT-HF assesses multiple primary endpoints in various domains, including LV structure (e.g., volumes), LV function, LV morphology (e.g., scar volume), functional capacity, quality of life, clinical outcome, and biomarkers (Table 7).

If the combination of MSCs and CPCs proves to be beneficial, the question will arise as to which cell type(s) is/are responsible for this effect. Answering this question requires a four-arm design, which makes it possible not only to ascribe to MSCs, CPCs, or both, any favorable changes observed in the MSCs + CPCs group, but also to determine whether the actions of MSCs and CPCs are additive or synergistic. An additional reason for including the CPCs alone group is to continue the evaluation of CPCs that was started in SCIPIO^{30, 31}. Although in that trial patients receiving CPCs exhibited an improvement in LV function, scar size, functional capacity, and quality of life^{30, 31}, SCIPIO was a small, open-label, first-in-human trial designed to assess feasibility and safety, not efficacy, in patients treated 3-4 months after bypass surgery. Therefore, it is important that the encouraging results of SCIPIO be rigorously tested in a Phase II study using a randomized, double-blind, placebo-controlled, multicenter design and without limiting the analysis to the post-bypass surgery setting. In addition, whereas in SCIPIO the route of administration was intracoronary^{30, 31}, CONCERT-HF will provide information regarding the efficacy of CPCs administered transendocardially.

Preclinical and clinical studies of MSCs in chronic ischemic HF

MSCs are a particularly promising cell type for cardiac reparative therapy because of their availability, immunomodulatory properties, and track-record of safety and efficacy⁴. In rodent and swine models of acute MI, administration of MSCs (both autologous^{8, 9} and allogeneic^{6, 7, 10, 11}) enhances recovery of LV function. In swine models of chronic MI, MSCs (both autologous¹² and allogeneic¹³), delivered either via epicardial injection¹² or transendocardial injection¹³), augment LV function^12–14. In both settings, MSCs have been found to promote angiogenesis, decrease collagen deposition, and reduce myocyte apoptosis⁴. Importantly, MSCs interact with endogenous precursor cells (i.e., CPCs) that may play an important role in cardiac repair⁷ (vide infra).

In the clinical arena, several studies have assessed the effects of MSCs, delivered via the transendocardial route, in patients with chronic ischemic HF^{4, 16, 18, 20–23, 40, 41}. The POSEIDON trial¹⁶ assessed the safety and efficacy of autologous and allogeneic MSCs (no placebo control group was studied). Compared with baseline, at 1 year both autologous and allogeneic MSCs reduced scar size, but only autologous MSCs improved the 6-min walk distance and MLHFQ score. Neither cell type increased LVEF. The TAC-HFT trial¹⁸ evaluated autologous MSCs and unfractionated mononuclear BM cells (BMCs). Compared with placebo at 1 year, both MSC and BMC administration improved the MLHFQ score, but only MSCs improved regional LV function and 6-min walk distance and reduced scar size (although they did not affect LVEF). Autologous MSCs incubated in vitro with a “cardiogenic” cocktail were found to have beneficial effects in the C-CURE trial⁴⁰ but not in the subsequent larger CHART-1 trial⁴¹, although in the latter a subgroup analysis suggested benefit in patients with LVEDV >200 ml. A dose-escalating study of allogeneic STRO-1^bright mesenchymal progenitor cells found a reduction in HF-related MACE with the highest dose (150×10⁶ cells) at 3 years, although there was no improvement in LVEF or LV volumes²¹. In all of these studies^{16, 18, 21, 40, 41}, administration of MSCs was safe, and allogeneic cells did not elicit a detectable immune reaction.

Preclinical and clinical studies of CPCs in chronic ischemic HF

The ability of CPCs to alleviate LV dysfunction has been repeatedly demonstrated by several independent laboratories in various preclinical animal models of acute MI¹. CPCs have also been shown to alleviate LV dysfunction and remodeling in rat^{26, 27, 29, 42} and porcine²⁸ models of chronic MI, where they reduced not only scar size but also collagen deposition in the noninfarcted region^{26, 27, 29, 42}. Because of the poor survival of transplanted cells, these effects have been ascribed to paracrine actions^{5, 43}. The feasibility and safety of CPC therapy in the clinical setting are supported by the SCIPIO trial, a Phase I, open-label trial of autologous CPCs in patients with chronic ischemic HF^{30, 31}. In SCIPIO, CPCs were isolated and expanded in all 20 treated patients, and their intracoronary (i.c.) infusion did not result in detectable adverse effects. At 2 years, CPC-treated patients exhibited improved LV function, functional status (NYHA class), and MLHFQ score as well as a decrease in scar size^{30, 31}. These results provide the rationale for studying the CPCs alone group in CONCERT-HF.

Preclinical studies of the MSCs + CPCs combination

The rationale for examining the combination of MSCs and CPCs stems from studies in porcine models in which administration of MSCs 3 days after MI induced proliferation and differentiation of endogenous CPCs⁷, suggesting that an MSC-CPC interaction could be an important mechanism for the beneficial effects of MSCs. A subsequent study demonstrated that the combination of both cell types (i.e., human MSCs and CPCs in immunosuppressed pigs), given transendocardially 14 days after MI, produced a greater reduction in scar size and improvement in LV function than each cell type alone, restoring diastolic and systolic LV function toward normal²⁴. Additive beneficial effects of MSCs and CPCs have also been observed in rats²⁵. Importantly, an additional porcine study was conducted, employing autologous porcine cell combination therapy in a model of chronic MI, with cells delivered by NOGA catheter⁴⁴. In this study, designed to recapitulate the delivery of cells that would occur in humans, the combination of MSCs and CPCs produced synergistic improvements in global and regional LV function relative to either cell type alone⁴⁴. Together, these findings illustrate potentially important biological interactions between MSCs and CPCs that enhance cell therapeutic responses.

Rationale for dose selection

The target dose of MSCs in CONCERT-HF (150×10⁶) was chosen on the basis of a dose-escalating trial of STRO-1^bright MPCs²¹ (cells similar to MSCs), in which increasing doses of MPCs were associated with decreasing LVEDV and LVESV; the benefits were maximal at a dose of 150×10⁶ cells, and no serious adverse events were reported. The safety of this dose is further supported by the fact that transendocardial injection of up to 200×10⁶ MSCs was safe in preclinical porcine models^{12, 13} and in the TAC-HFT¹⁸ and POSEIDON trials¹⁶, and that intravenous doses of MSCs >200×10⁶ cells have been used with no significant adverse effects¹⁵.

The choice of the target dose of CPCs (5×10⁶) reflects a number of considerations. Pig studies have shown that post-MI LV dysfunction is effectively ameliorated by 5×10⁵ CPCs (a dose equivalent to ~1×10⁶ CPCs in humans) given by the i.c. route²⁸ or 1×10⁶ CPCs (equivalent to ~2×10⁶ CPCs in humans) given transendocardially²⁴. In the only clinical trial of CPCs to date, SCIPIO^{30, 31}, 1×10⁶ CPCs were given by the i.c. route, with encouraging results regarding LV function, scar size, functional status, and quality of life³⁰. Furthermore, i.c. infusion of higher CPC doses in pigs (up to 20×10⁶, equivalent to ~40×10⁶ CPCs in humans) is well tolerated and produces no discernible adverse effects, as assessed by cardiac enzymes, myocardial function, and histology⁴⁵. Thus, both preclinical and clinical evidence supports the notion that 5×10⁶ CPCs should be effective in humans. Although higher doses could conceivably have greater efficacy, it would be challenging to reproducibly generate >5×10⁶ CPCs from small amounts of cardiac tissue (10-20 mg). In conclusion, the target cell doses for CONCERT-HF (150×10⁶ MSCs and 5×10⁶ CPCs) have a strong safety profile and there is considerable evidence for their potential to improve LV function and functional status in patients with HF.

Rationale for using the transendocardial route of administration

In previous studies in porcine models^{7, 10, 11, 13, 24} and in humans^{16, 18, 21, 40, 41}, transendocardial delivery of MSCs improved post-MI LV dysfunction and remodeling and (in clinical trials) functional capacity and MLHFQ score. The efficacy of intramyocardial injections of CPCs has been well documented in rodents^{25, 46} and in pigs^{1, 24}. Although the only previous clinical trial of CPCs used i.c. infusion^{30, 31}, we chose the transendocardial route for CPCs so that MSCs and CPCs could be delivered together. Transendocardial delivery was the route used for both cell types in the porcine studies that support the superiority of the combination over either cell type alone^{7, 24}; presumably, the juxtaposition of MSCs and CPCs promotes their interactions and may be the basis for their additive therapeutic effects⁴. Additional advantages of transendocardial delivery are that it is associated with greater myocardial retention⁴⁷ and makes it possible to reach regions distal to occluded coronary arteries.

Use of MRI in patients with cardiac devices

MRI was selected over other imaging modalities, such as cardiac CT and echocardiography, because it enables accurate, high-resolution assessment of myocardial scar^48-50, is considered to be more reproducible for measuring LV function and volumes^{51, 52}, and does not require any radiation exposure. One of the innovative aspects of CONCERT-HF is that MRI is used in patients with pacemakers and/or implanted defibrillator devices. Traditionally, these patients have been excluded because of the concern that the MRI exam may interfere with the function of the device and also because of the suboptimal image quality^{53, 54}. However, studies have shown that with appropriate study candidate selection as well as software, equipment, and positioning optimization, MRI scanning can be performed safely in the presence of most cardiac devices without precluding acquisition of meaningful data^{33, 34}. In this trial, extensive in-person and webinar-based training procedures were put in place at the local Radiology and Cardiology departments at each site under the supervision of the MRI Core lab to ensure patient safety and proper execution of the MRI protocol. The initial experience garnered at the MRI Core at Johns Hopkins University shows that, with the use of the exclusion criteria listed in Table 2 and appropriate software^55-57, MRI scans can be obtained safely and the images can be analyzed for LV volumes and scar size in the vast majority of the cases. The experience of CONCERT-HF will generate new insights into the feasibility of using of MRI in patients with cardiac devices, which may have an important impact on the design of future studies not only of cell therapy but also of other interventions in HF. This information also may help obviate the problems stemming from the fact that patients may acquire devices over the course of a trial, which results in significant drop-off of patients when serial MRI imaging is utilized to assess changes in LV function over time⁵⁸.

Limitations

Because of the innovative nature of CONCERT-HF, the feasibility of certain aspects of the study is unclear. EMBs may not be sufficient to generate the target dose of CPCs (5×10⁶) in all patients. The long interval between EMB and product injection (necessitated by the time needed for expansion) may cause some patient drop-out. Since this is the first trial to use MRI in patients with devices, the quality and reproducibility of images are not known. CONCERT-HF will furnish important feasibility information regarding these issues. An additional possible limitation is that, depending on variability, the trial may not have enough power to assess some of its endpoints.

In conclusion, CONCERT-HF is the first cell therapy clinical trial to evaluate the combination of two cell types. This study will provide important insights into the potential therapeutic utility of MSCs and CPCs, given alone and in combination, for patients with ischemic HF. In addition, the collection and storage of HF patient-specific biomaterials may offer important tools for future ancillary studies of regenerative medicine.

NOVELTY AND SIGNIFICANCE.

What Is Known?

• Bone marrow mesenchymal stem cells (MSCs) ameliorate left ventricular (LV) dysfunction and reduce scar size in preclinical models of chronic ischemic cardiomyopathy (ICM), and several clinical trials suggest that MSCs are safe and beneficial in patients with chronic heart failure (HF) of ischemic etiology.

• c-kit+ cardiac progenitor cells (CPCs) also attenuate LV dysfunction and remodeling in animal models of chronic ICM. The SCIPIO trial has suggested benefit in patients with ischemic HF, but the study was not designed to establish efficacy.

• When MSCs and CPCs are combined in preclinical models, the therapeutic effects are additive, and thus superior to those of either cell type alone. However, neither the concept of combinatorial stem cell therapy nor the relative efficacy of MSCs vs. CPCs has been tested in humans.

What New Information Does This Article Provide?

• We describe CONCERT-HF, a Phase II trial that assesses the feasibility, safety, and efficacy of autologous MSCs and CPCs, given alone or in combination, in patients with ischemic HF.

• CONCERT-HF is the first trial to evaluate the effects of CPCs in ischemic HF with a multicenter, randomized, double-blind protocol powered for efficacy.

• CONCERT-HF is the first trial to compare the relative effects of MSCs and CPCs in patients in the same study.

• CONCERT-HF is the first trial to evaluate a combination of two different cell types.

• Other innovative features of CONCERT-HF include the adaptive design and the use of magnetic resonance imaging (MRI) in patients with devices.

The four-arm design (MSCs alone, CPCs alone, MSCs + CPCs, and vehicle) enables CONCERT-HF to address multiple questions: i) Is combined treatment with autologous MSCs and CPCs feasible and safe in patients with ischemic HF? ii) Does administration of MSCs alone have beneficial effects on LV dysfunction and remodeling, scar size, functional capacity, quality of life, and/or clinical outcome? iii) Does administration of CPCs alone have beneficial effects on these variables? iv) Does administration of MSCs + CPCs have beneficial effects on these variables? v) Is the combination of MSCs and CPCs superior to either cell type alone? vi) When compared to each other, is either cell type alone superior? Each of these questions has obvious significance for the development of cell-based therapies in patients with ischemic HF. In addition, the collection and storage of HF patient-specific biomaterials may offer important tools for future ancillary studies of regenerative medicine.

Nonstandard Abbreviations and Acronyms

BM

bone marrow

BMA

bone marrow aspiration

BMCs

bone marrow cells

BRC

CCTRN Biorepository

CCMF

Central Cell Manufacturing Facility

CPCs

c-kit⁺ cardiac progenitor cells

DCC

Data Coordinating Center

EMB

endocardial biopsy

HF

heart failure

HF-MACE

major adverse cardiac events related to heart failure i.c. intracoronary

ICM

ischemic cardiomyopathy

LV

left ventricular

LVEDV

left ventricular end diastolic volume

LVEF

left ventricular ejection fractions

LVESV

left ventricular end systolic volume

MI

myocardial infarction

MLHFQ

Minnesota Living with Heart Failure Questionnaire

MNCs

mononuclear cells

MRI

magnetic resonance imaging

MSCs

mesenchymal stem cells

QC

quality control

RHC

right heart catheterization

SD

standard deviation

VO₂ max

maximal oxygen consumption