Treatment with bone marrow-derived mononuclear cells (BM-MNC) may improve left ventricular (LV) function in patients with chronic ischemic heart disease (IHD). Delivery method of the cell product may be crucial for efficacy.
Treatment with bone marrow-derived mononuclear cells (BM-MNC) may improve left ventricular (LV) function in patients with chronic ischemic heart disease (IHD). Delivery method of the cell product may be crucial for efficacy.
We aimed to demonstrate that the combination of intramyocardial and intracoronary injection of BM-MNC is safe and improves LV function in patients with chronic IHD.
After a safety/feasibility phase of 10 patients, 54 patients will be randomly assigned in a 1:1:1 pattern to 1 control and 2 BM-MNC treatment groups. The control group will be treated with state-of-the-art medical management. The treatment groups will receive either exclusively intramyocardial injection or a combination of intramyocardial and intracoronary injection of autologous BM-MNC. Left ventricular function as well as scar size, transmural extension, and regional wall-motion score will be assessed by cardiac magnetic resonance imaging studies at baseline and after 6 months. The primary endpoint is the change in global LV ejection fraction by cardiac magnetic resonance from 6 months to baseline.
The results, it is hoped, will have important clinical impact and provide essential information to improve the design of future regenerative-medicine protocols in cardiology.
As cell delivery may play an important role in chronic IHD, we aim to demonstrate feasibility and efficacy of a combined cell-delivery approach in patients with decreased LV function.
Primary percutaneous coronary intervention significantly reduces early mortality and improves late clinical outcome in patients with acute myocardial infarction (MI). However, congestive heart failure secondary to adverse ventricular remodeling following MI continues to be a major medical problem worldwide.[2, 3] Among different potential strategies that could prevent these adverse processes, cell-based regenerative therapies may play a promising role.
Injection of bone marrow-derived mononuclear cells (BM-MNC) has been proposed as a new therapeutic perspective for patients with chronic ischemic heart disease (IHD). Application of the cell product is possible in different ways: BM-MNC may be injected into the coronary artery, corresponding to the territory presenting ischemia, if the vessel is still open. Furthermore, the cells may be injected directly into the ischemic myocardium of the left ventricle (LV) from the endocardium using special injection catheters, or from the epicardium during coronary artery bypass graft surgery. Initially, smaller studies demonstrated that intramyocardial BM-MNC injection in patients with chronic myocardial ischemia is safe.[4, 5] A beneficial effect on myocardial perfusion and LV function has been shown by some, but not all, randomized, double-blind, placebo-controlled trials.[6, 7] As intracoronary injection of BM-MNC also has been shown to have some benefit in chronic IHD, a combined approach has been proposed.
A subset of patients with chronic IHD presenting with refractory angina received particular benefit from bone marrow-derived cell therapy, significantly reducing anginal complaints and improving exercise capacity.
The aim of the presented trial is to confirm feasibility and safety of a combination of intracoronary and intramyocardial injection of BM-MNC in patients with chronic IHD in the ongoing safety phase. In a second multicenter randomization phase, the trial will be expanded, maintaining a single, centralized cell-processing facility. The objective of the randomization phase is to confirm beneficial effects of such therapy in terms of an improvement in LV function, assessed by cardiac magnetic resonance (CMR) imaging.
The study protocol was approved by the ethical committee of the Canton of Ticino as well as by the federal competent authorities (Swissmedic and Bundesamt für Gesundheitswesen).
As demonstrated by the study-design flow chart (Figure 1), the study consists of 2 phases: phase I, defined as a safety and feasibility phase; and phase II, a randomization phase. Phase I is a monocenter study including 10 patients with chronic IHD (defined as ≥1 documented previous MI or due to documented ongoing silent ischemia without any option for revascularization) and decreased left ventricular ejection function (LVEF; ≤40%, as assessed by transthoracic echocardiography) in an open-label, noncontrolled manner. Patients fulfilling the inclusion criteria and none of the exclusion criteria (Tables 1 and 2) are requested to provide their informed consent to participate in the study. They will be treated, in addition to receiving best medical management (Table 3), by intramyocardial injection of BM-MNC, the first 5 patients exclusively and the following 5 patients additionally receiving intracoronary injection of BM-MNC, dividing the cell product into 2 equal portions.
|Chronic ischemic heart disease; LVEF ≤40% in a stable phase of the disease without option for further revascularization|
|>3 months after last MI or last revascularization (PCI or CABG)|
|Patency of the infarct-related artery, presumably responsible for the most extensive MI|
|NYHA class II–IV or CCS class II–IV (≥ class III according to 1 of the 2 classifications)|
|Age >18 years|
|LV dysfunction of nonischemic origin|
|Moderate to severe aortic valve disease; aortic or mitral prosthetic valve|
|Known active infection or chronic infection with HIV, HBV, or HCV|
|Significant mitral valve insufficiency (ERO >0.2 cm2 with possibility of mitral valve surgery)|
|LV thrombus on echocardiography or CMR|
|LV wall thickness <5 mm in the target territory|
|Congenital heart disorder of hemodynamic relevance|
|Chronic inflammatory disease|
|Serious concomitant disease with a life expectancy <1 year|
|Follow-up unlikely (eg, no fixed residence)|
|Contraindication for CMR imaging (ie, pacemaker, neurostimulator, severe claustrophobia)|
|Severe renal failure (Cr >250 mmol/L or GFR <30 mL/min/1.73 m2)|
|Relevant liver disease (GOT >2× normal or spontaneous INR >1.5)|
|Anemia (Hb <8.5 mg/dL), thrombocytopenia (<100 000/μL)|
|Participation in a clinical trial within the last 30 days|
|ACEI or ARB, highest dose tolerated unless contraindication|
|β-Blocker, highest dose tolerated unless contraindication|
|Spironolactone or eplerenone, provided the patient is NYHA class III|
|Medical therapy for CAD including ASA, clopidogrel/prasugrel/ticagrelor, and statins|
In phase II, patients fulfilling the same inclusion criteria and none of the exclusion criteria as described above are randomized, using closed envelopes in a 1:1:1 manner, into 2 treatment groups depending on whether BM-MNC are applied (on top of best medical management) exclusively by intramyocardial injection or in combination with intracoronary injection or control. Patients in the control group will receive best medical treatment (ie, state-of-the-art antiplatelet regimen, statin, β-blocker, spironolactone or eplerenone, and an angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker unless contraindicated; Table 3). After 6 months of follow-up, all patients in the control group will then be further randomized into the 2 treatment arms, as described above.
All 3 groups will undergo serial CMR studies at baseline and after 6 months. Patients initially randomized into the control group will be studied for 12 months (ie, 6 months after BM-MNC treatment). Due to the crossover design of the study, no 12-month follow-up will be available for the nontreated control patients. As the study is not powered to detect differences in clinical events between the groups, it was thought better to allow the nontreated control patients to undergo cell therapy themselves after 6 months, rather than insisting on long-term follow-up.
The primary objective of phase I of the study is to determine the safety and feasibility of the technique, especially of the combination of intramyocardial and intracoronary administration of BM-MNC in patients with chronically decreased LVEF due to IHD. Thus, safety endpoints consist of serious adverse events of any type during hospitalization and after 30 days, such as death, bleeding, MI, stroke, or cardiac reintervention of any type. Furthermore, troponin levels will be assessed before and after the procedure.
The primary endpoint of phase II is the change in global LVEF by CMR at 6 months compared with baseline. Comparisons will be made between each treatment modality and control. The secondary end points are:
Patients will undergo bone marrow aspiration within 24 hours prior to cell administration. For the latter, 110 mL of bone marrow will be collected from the iliac crest under local anesthesia, diluted in 1000 IU/10 mL heparin and transferred at room temperature (20 °C ± 5 °C) to the Good Manufacturing Practices cell-processing facility (Cell Therapy Unit of Ticino Cardiac Center). Serology must be negative for hepatitis B surface antigen (HBsAg) and for anti-human immunodeficiency virus (HIV) 1/2 and anti-hepatitis C virus (HCV) antibodies.
The mononuclear cell fraction is isolated according to a standard protocol, with minor modifications. Briefly, the bone marrow is prefiltered and subjected to density gradient centrifugation on the Ficoll-Paque Premium centrifugation medium (GE Healthcare, Uppsala, Sweden). After several washings, isolated mononuclear cells are suspended in 5% human albumin solution (Albumin CSL 5%, CSL Behring AG, Bern Switzerland; CSL Behring AG, Bern, Switzerland) for intravenous infusion and filled into syringes.
Samples are collected for quality-control analyses, which include cell viability (7-aminoactinomycin D uptake) and immunophenotype (CD45/CD34/CD133) by flow cytometry (FC 500; Beckman Coulter, Fullerton, CA); determination of white blood cell concentration, percentage of lymphocytes/monocytes/granulocytes, platelet concentration, and hematocrit by automated cell counter (ABX Micros 60; Horiba Medical, Irvine, CA); bacterial endotoxins by chromogenic limulus amebocyte lysate test (Endosafe-PTS; Charles River Laboratories, Wilmington, MA); and sterility by microbiological control for cellular products (European Pharmacopoeia 2.6.27) using an automated microbial detection system (BacT/ALERT 3D; bioMérieux Clinical Diagnostics, Marcy l'Etoile, France). All the analyses are completed in few hours, with the exception of the microbiological control of cellular product, which requires 7 days.
The product (50–500 × 106 fresh cells in 9–18 mL 5% human albumin) is stored in quarantine at 10 °C ± 5 °C until the preinfusion release assays are completed. If the preinfusion results comply with specifications (cell viability ≥80%, lymphocytes ≥25%, monocytes ≥4%, granulocytes ≤75%, total platelets ≤2 × 109, hematocrit ≤3%, endotoxin <5 EU/mL), the product is administered to patients within 24 hours.
Vascular access is gained via introduction of an 8-Fr sheath in the femoral artery. After a conventional biplane LV angiography and administration of heparin for an activated clotting time of 200 to 250 seconds, a mapping catheter (NOGA XP Cardiac Navigation System; Cordis Corp) is placed in the LV and electromechanical mapping is performed as previously described. The target zone for BM-MNC injection is then selected first by avoiding areas with transmural or near transmural myocardial scar tissue, as expressed by CMR delayed enhancement. Provided that areas of myocardial ischemia in the previously performed perfusion imaging (CMR) are present, those areas will be further identified via electromechanical mapping, searching for locations with viable myocardium (unipolar voltage >6.9 mV, preferably between 7 and 12 mV) and associated decreased mechanical activity, indicating hibernating but viable myocardium. In the absence of myocardial ischemia at CMR, the border zone of scar shall be targeted for injection.
A steerable injection catheter (Myostar; Biosense Webster, Diamond Bar, CA) is then placed in the LV and the tip guided manually to the previously determined target region (Figure 2). Zones with electrical activity between 7 and 12 mV and with mechanical activity are preferred; zones with electrical activity <7 mV and without mechanical activity shall be avoided. Prior to injection, stability of the catheter on the endocardial surface must be established. The catheter tip shall be in perpendicular position against the ventricular wall with loop stability <3 mm.
Once the appropriate injection site has been identified, the needle is extended into the myocardial wall, which can be verified by the presence of premature beats on the intracardiac electrocardiography at the moment of the needle protrusion. At that point, 0.5 mL of the BM-MNC solution is injected over a period of 1 minute using a 1-mL Luer-Lock syringe (Medline Industries, Mundelein, IL). After retraction of the needle, a new position is evaluated for injection and the procedure is repeated. In absence of ischemia or hemodynamic complications, 18 injections are made. Injections should not be targeted to the mitral valve annulus and the apex of the heart, as injections in this area could result in cardiac rupture or pericardial effusion. Furthermore, injections will not be done if the wall thickness of the target area is <5 mm as assessed by CMR, to minimize the risk of rupture of the LV wall.
Via the 8-Fr sheath in the femoral artery, a coronary guiding catheter will be placed in the ostium of the former infarct-related vessel. Cells will be infused via an over-the-wire balloon catheter advanced into the infarct-related coronary artery, if present at the level of a previously implanted coronary stent. The balloon is then inflated with low pressure (2–4 bar) and 3 mL of the BM-MNC cell suspension is administered. Blood flow remains completely blocked for 3 minutes to allow for adhesion and transmigration of the infused cells through the endothelium. This maneuver will be repeated 3 times to allow infusion of the entire 9-mL progenitor cell suspension, interrupted by 5 minutes of reflow by deflating the balloon to minimize extensive ischemia. After completion of intracoronary cell reinfusion, coronary angiography will be repeated to ascertain vessel patency, absence of embolization, and unimpeded flow of contrast material.
Cell administration will be performed under the standard antithrombotic therapy for coronary interventions (ASA, heparin, and prasugrel, ticagrelor, or clopidogrel). Periprocedural safety of the BM-MNC infusion will be monitored by assessment of serum cardiac enzymes including troponin on the day after the intervention. Periprocedural MI is defined as previously described.
Patients will undergo CMR studies at baseline (ie, some days before undergoing NOGA mapping and cell administration), as well as 6 months after cell therapy. Patients in the control group will undergo 3 CMR studies, because after the 6-month follow-up they will be further randomized into the 2 treatment arms following the crossover design of the study and will therefore undergo a third CMR study 12 months after inclusion in the study (ie, 6 months after cell therapy). All patients will be examined in supine position on a 3.0-Tesla magnet (Magnetom Skyra; Siemens, Erlangen, Germany) using a 36-channel cardiac coil. Patients will be instructed to refrain from caffeine and medications such as aminophylline for 24 hours before CMR. Following localizer acquisitions, cine electrocardiography-triggered standard steady-state free precession images will be acquired in the 2-, 3-, and 4-chamber orientations, as well as in a stack of contiguous short-axis slices covering the entire LV for functional analysis.
Adenosine stress CMR first-pass perfusion imaging will be performed after a 3-minute intravenous adenosine infusion (140 µg/kg/min) and a peripheral gadolinium bolus (gadobutrol; 0.1 mmol/kg body weight) in 4 evenly spaced short-axis planes covering the entire LV from base to apex by using a saturation-recovery gradient-echo sequence.[16-18]
After stress perfusion an additional bolus of gadolinium (gadobutrol; 0.1 mmol/kg body weight) will be given for delayed gadolinium enhancement imaging (viability/scar imaging). Finally, 10 to 15 minutes after contrast application, a 2D phase-sensitive inversion recovery gradient-echo sequence will be obtained for delayed gadolinium enhancement in 3 long-axis orientations and in a contiguous stack of short-axis images by using the same plane directions as for cine imaging. Cardiac magnetic resonance data analysis will be performed using a dedicated cardiac analysis software (CMR42, version 3.4.0; Circle Cardiovascular Imaging, Calgary, Canada). Left ventricular end-diastolic and end-systolic volumes, LVEF, and LV mass will be quantified based on the contiguous short-axis slices of cine images. For functional analysis, systolic thickening will be graded as normal (score 0), mildly hypokinetic (1), severely hypokinetic (2), akinetic (3), and dyskinetic (4) by visual analysis and by using a 17-segment model.
Adenosine stress CMR first-pass perfusion imagines will be interpreted in a segmental analysis (exclusion of segment 17). Myocardial ischemia will be defined as a stress-induced myocardial hypoenhancement in absence of hyperenhancement at delayed gadolinium images (scar) in the same location.
On delayed gadolinium enhancement images, myocardial infarct size will be quantified using the full-width at half-maximum method. Transmurality of the scar will also be visualized and described in a segmental analysis. Scar mass will be expressed in grams and as a percentage of the total LV mass. Changes over time (baseline, 6 months) of volumetric, functional, scar, and perfusional parameters will be compared.
Whereas CMR images are already taken into account to identify the target zone for intramyocardial injection of BM-MNC, all analyses concerning the primary and secondary efficacy endpoints of phase II of the study will be entirely performed at the end of the trial (ie, after completion of the last CMR). Two to 3 experienced operators from a core lab—situated at Ticino Cardiac Canter in Lugano, blinded to randomization status, and without any contact to the individual patients—will perform the analyses following a standardized definition. To assess interoperator variability, in 10% of patients CMR analyses will be performed by 2 independent operators. Entering phase II of the trial, each center that is participating in the trial will receive adequate teaching about the standardized CMR protocol before including the first patient.
Patients are considered entered into the trial and included in the analysis of safety in terms of an intention-to-treat analysis when informed consent has been signed and the patient is randomized. Descriptive statistics (such as minimum, median, mean, maximum, and SD) of continuous variables will be presented. Nominal variables are summarized in terms of frequencies and percentages.
Entering the randomization phase of the trial, the statistical comparisons of both treatment arms with control as for the primary and the secondary endpoints are performed using analysis of covariance (ANCOVA). For the 2 primary endpoint comparisons, namely of both experimental arms to control, a Bonferroni corrected P value ≤0.025 is considered significant. Suitable 95% confidence intervals for both differences of LVEF changes between treatment arms will be provided.
The influence of the total number and the number of CD34+ and CD34/CD133+ cells of infused progenitor cells on LVEF changes and clinical outcome is analyzed by means of regression analyses with appropriate interpretive terms.
For the analysis of the secondary endpoints, all P values ≤0.05 will be considered significant. As multiple testing will be performed, we will consider all results “exploratory.” For the analysis of binary endpoints, comparisons will be performed using χ2 or Fisher exact test, depending on the expected cell frequencies. For continuous outcomes, an independent samples Student t test or Wilcoxon test is used.
Major adverse cardiac event is defined as the occurrence of one of the predefined clinical scenarios (death, MI, coronary revascularization, rehospitalization for heart failure, stroke) and will be assessed and compared between groups at 6, 12, and 24 months. Because at the time of analysis not all patients will have suffered from a MACE, the secondary endpoint time from randomization to MACE may be censored for these patients. Kaplan-Meier estimates of survival curves will be shown and median time to MACE will be computed, including 95% confidence intervals according to the method of Brookmeyer and Crowley. Estimated survival curves will be compared between treatment groups using a log-rank test.
The study is powered for the primary endpoint of the change in LVEF determined by CMR at 4 months compared with baseline. As patients are in the stable phase of chronic ischemic disease, we do not expect any change in LVEF for the control group. For both treatment groups we assume an improvement in global LVEF as assessed by quantitative CMR of 3.0% to 3.5% (according to previous studies).[7, 8] In these stable patients we do not expect an SD >3%. For a value α = 0.025 and a power of 80%, we therefore need a sample size between 14 and 20 per group. We thus decided to include 18 patients per group, resulting in a final sample size for the control group of n = 18 and for both treatment groups of n = 27.
Local clinical-trial units will assure adequate data management following the rules of good clinical practice, and regular monitoring will be performed. After completion of phase I of the trial, an independent data safety monitoring board will review all cases and all adverse events. Results will then be presented to the federal competent authorities (Swissmedic).
Progenitor cell-based therapy is among the most promising new strategies to reduce initial cardiac damage and the deleterious LV remodeling after acute MI as well as to repair cardiac damage once it occurs. Therapy with BM-MNC is thought to have the capacity to induce neoangiogenesis in a paracrine manner and may therefore figure as a potential source in regenerative medicine. For this reason, their role in chronic ischemic cardiomyopathy cannot be the regeneration of new myocardium or the transformation of scar tissue in functionally contracting tissue. The potential target of BM-MNC is, therefore, the peri-infarct zone, where ischemia at least to some degree may be present in most of the patients. Injected either in the infarct-related artery or directly into the ischemic myocardium, BM-MNC may enhance neovascularization and therefore improve LV function.
Different cell-application techniques may have an impact on efficacy of the therapy. Schächinger et al nicely demonstrated that the homing capacity of intracoronary-injected progenitor cells is highest in the acute phase after MI, with a subsequent diminution the longer after MI that the treatment is administered. This could be due to a subsequent diminution of the microvascular bed in the peri-infarction zone, to a diminished attraction of the progenitor cells by the less-freshly infarcted myocardium, as well as to a diminished homing capacity of the cells. For this reason, most clinical trials in the chronic phase of IHD have been done with direct intramyocardial injection of the progenitor cells in the border zone of the MI. The disadvantage of this technique, apart from the higher risk of complications, is the technical difficulty of targeting exactly the area of interest to apply the cell product in a homogenous manner. Together with CMR delayed enhancement imaging and perfusion imaging, electromechanical mapping of the LV using the NOGA system may help to define and engage target zones of ischemic, but viable, myocardium, or, in absence of overt ischemia, the border zone of myocardial scar.
Here we present and discuss the rationale and study design of a multicenter, randomized, controlled protocol to evaluate the benefit of BM-MNC transplantation compared with standard therapy in patients with chronic ischemic disease and reduced LV function. The goal of the study is initially to demonstrate feasibility and safety of the combination of intramyocardial and intracoronary application of BM-MNC. Further, we aim to confirm efficacy by hypothesizing a significant increase in LVEF in the treated patients. We therefore designed a trial with an open-labeled, monocenter feasibility phase, followed by a multicenter phase with 2 BM-MNC treatment groups. One group will receive intramyocardial administration of BM-MNC exclusively, and the other group will receive the BM-MNC via a combination of intramyocardial and intracoronary administration.
The combined use of CMR and electromechanical mapping of the LV will allow evaluation of ischemic but viable zones of the LV in detail, and thus will define and target zones for BM-MNC treatment. The results, it is hoped, will have important clinical impact and provide essential information to improve the design of future regenerative-medicine protocols in cardiology.
As a potential limitation of the study, it has to be pointed out that the tested treatment regimen consists of unselected mononuclear cells, and therefore reflects only one aspect of a growing variety of different cell-therapy modalities. Future studies have to show if stimulation or “treatment” of the bone marrow may enhance function or density of the BM-MNC or whether other cell types—such as selected mononuclear cells, cardiac stem cells, induced pluripotent stem cells, or allogenic cells from healthy donors—will exert stronger effects on myocardial regeneration.
Our choice to assess LV function with CMR allows us to obtain detailed information in terms of regional LV function, scar size, and perfusion. At the same time, patients with severely reduced LV function who are treated following guidelines with an implantable cardioverter-defibrillator or cardiac resynchronization cannot be included in the trial, which must be considered a further important limitation of the study.