Peripheral intravenous infusion of stem cells is an attractive noninvasive method of stem cell delivery due to its simplicity and applicability. Human bone marrow cells have been shown to “home” to peri-infarct areas when infused into the peripheral circulation of a mouse model of acute myocardial infarct [45]. However, this technique may not be suitable in the chronic heart failure setting, which lacks the temporarily upregulated biological homing signals present in the AMI setting. Another significant limitation is that only a few cells appear to reach the affected area due to trapping of the cells in the microvasculature of the lungs, liver, and lymphoid tissues [46]. Peripheral infusion has been extensively used in animal models, but there has been only one human study to date – a placebo-controlled study assessing primarily the safety of peripheral infusion of allogeneic human mesenchymal stem cells (Prochymal™) after acute myocardial infarction [47]. This was confirmed as a safe delivery method, and the treatment group showed significant improvement in symptoms, LV ejection fraction, as well as evidence of reverse remodelling on cardiac MRI. However, given the already small numbers of injected cells that are known to be retained by the heart, it is likely that more direct interventional injection will be more efficacious.

Cell mobilization provides an indirect method of introducing cell therapy to the injured myocardium. In the clinical setting, one of the most studied mobilizing factors is granulocyte-stimulating factor (G-CSF). Endogenous G-CSF is a potent hematopoietic cytokine that affects progenitor cell proliferation, maturation, and functional activation and enhances the mobilization and recruitment of stem cells from the bone marrow to the blood circulation [48]. G-CSF is also proposed to have a range of direct effects on the myocardium, including creating a favorable environment for homing and engraftment of stem cells, inhibiting apoptosis, and promoting cardiomyocyte survival, including angiogenesis and reducing fibrosis associated with adverse cardiac remodelling [49]. G-CSF is generally well tolerated with most common side effects being bone pain and fevers.

Initial preclinical animal studies [50] and phase 1 clinical trials after myocardial infarction showed that G-CSF treatment appeared to be safe with possible beneficial effects on left ventricular systolic function [51, 52]. However, a subsequent randomized placebo-controlled study of subcutaneous G-CSF after primary PCI for AMI showed no additional improvements in left ventricular function when compared to placebo [53]. This was confirmed in a meta-analysis on the effect of G-CSF in AMI which showed no overall benefit although subset analysis showed there may be benefit limited to patients with LV systolic dysfunction and if the infusion is started early [54]. There has also been concerns raised regarding the potential of G-CSF to cause in-stent restenosis (ISR) as it increases the level of circulating neutrophils [55] which may accelerate the process of neointimal proliferation. One of the initial clinical trials reported an unexpectedly high rate of ISR in ten patients treated with G-CSF following an AM [56]. However, a more recent trial involving patients with large infarcts and late revascularization did not show any increased incidence of ISR [57]. Furthermore, an intracoronary intravascular ultrasound (IVUS)-based study, 5 months after stent insertion for AMI, showed no increase in in-stent neointimal hyperplasia in the G-CSF treatment group compared to placebo [58]. Two small, nonrandomized, studies have shown possible beneficial effects of G-CSF in chronic ischemic heart failure with regard to patient symptoms and improvement in LV function but also raised concerns regarding worsening of angina during the treatment phase [59, 60]. It remains to be determined if G-CSF treatment could be an effective part of a treatment strategy combining several cytokines and/or local stem cell delivery, and there are ongoing trials of its use (e.g., REGENERATE-IHD [61]) which will provide further data regarding this.

Retrograde infusion of stem cells through the coronary sinus and coronary venous system has been achieved in experimental models which suggest that this method may provide a more uniform delivery of cells with improved cell retention rates [31]. There have been sporadic case reports [62] demonstrating safety of this technique, but there have not been any published clinical trials of this technique. In this procedure, the coronary sinus is cannulated using a guidewire, and an angioplasty balloon is advanced to the selected cardiac vein related to the coronary artery territory to be treated. The venous blood flow is then halted, by prolonged balloon inflation for approximately 15 min, while the stem cells are infused under pressure. There is yet to be any clinical efficacy data regarding this technique, and the main limitations to this approach are the technical difficulties related to the more variable coronary venous anatomy, which can at times be difficult to negotiate with guidewires and balloons. There are also concerns relating to the potential myocardial injury caused by high-pressure retrograde infusion. This technique may not be possible in patients who already have a coronary sinus device in situ such as a LV lead for resynchronization therapy or a coronary sinus annuloplasty device although cells could be administered as an adjunct to the implantation of such devices.

Intracoronary infusion is the method that has been most widely adapted in clinical trials of cell therapy, especially after primary angioplasty for AMI. This method has also been used in patients with heart failure secondary to ischemic heart disease, although in this setting occluded coronary arteries may prevent targeted delivery to the infarcted area. The technique employs the skills of balloon angioplasty and therefore can be performed by an interventional cardiologist with minimal additional training. Cells are infused through the central lumen of an over-the-wire (OTW) balloon catheter during low-pressure balloon inflation for 2–4 min (“stop-flow” technique). Standard short (6–10 mm) balloons are used to minimize endothelial injury, and balloons are also undersized by 0.5 mm less than the reference vessel diameter. Balloon inflation time is empirically chosen, and the purpose is to prolong the contact time between the infused cells and endothelium of the coronary macro-/microcirculation as well as to provide an ischemic stimulus which may improve homing of cells to the target myocardium. Inflation time may need to be curtailed in patients who do not tolerate prolonged occlusion of a clinically important coronary artery. In the AMI setting infusion of cells is limited to the infarct-related artery, while in patients with chronic heart failure, cell infusion may be distributed amongst all patent coronary artery and graft conduits. Some more recent animal work suggests that single bolus infusion may be as effective [63] although stop flow seems to be the prominent method in the published trials.

A preclinical study in dogs raised concerns regarding the possibility of microinfarcts caused by intracoronary injection of mesenchymal stromal cells [64], and there is likely to be a safety threshold with regard to the size and dose of cells delivered using the intracoronary route. Reassuringly, the relative safety of this technique has been confirmed in multiple clinical trials, and a recent meta-analysis of 811 patients confirmed that the relative risks of mortality and morbidity, measured by incidence of reinfarction, arrhythmias, restenosis, hospital readmission, and target vessel revascularization, were not significantly increased in participants who received intracoronary BMSCs following AMI compared with controls [65].

Intramyocardial injection is the most direct method of administering cell therapy, and there are several advantages of this technique in patients with heart failure and angina. In contrast to the AMI setting, patients with chronic ischemic cardiomyopathy are unlikely to release signals from damaged myocardium to induce stem cell homing, and theoretically it may be more effective to use intramyocardial injection to deliver the cells to the target area. Also, in the chronic setting the coronary artery subtending the infarct/ischemic area may be occluded without significant collateral formation, and thus intramyocardial delivery provides a means for targeting these areas. Furthermore, experimental data using radionucleotide-labelled BMSCs suggest that target area cell retention may be higher with intramyocardial injection when compared to the intracoronary delivery method [66]. Experience of the intramyocardial route has been mainly limited to chronic conditions due to concerns of myocardial perforation and arrhythmia in the acute setting.

There are currently three methods for intramyocardial injection of cells: open surgical transepicardial injection, percutaneous transendocardial injection, and trans-coronary-venous intramyocardial injection of cells. Earlier clinical trials have relied on surgical transepicardial injection of cells where results are confounded by simultaneous coronary bypass surgery [34, 67–69]. When bypass is not indicated, percutaneous delivery may provide comparable efficacy with reduced risk. Studies in swine models of chronic ischemic heart failure comparing percutaneous and surgical transplantation of skeletal myoblasts have demonstrated that both approaches lead to similar improvements in LVEF and reverse remodelling [70].

Percutaneous transendocardial injection is usually performed via 8Fr femoral access although brachial access for transendocardial injection of stem cells has been reported [71]. A percutaneous approach allows this therapy to be delivered to the higher-risk population and also provides an option for repeated therapy if required. The main additional risks of this approach are ventricular perforation leading to pericardial effusion/tamponade and procedure-related ventricular arrhythmias although the reported incidences of these have been low in the published clinical trials. A deflectable tip injection needle catheter is advanced retrogradely across the aortic valve, and cells can be injected directly into any area of the left ventricular endocardium. There are four commercially available catheter systems currently being used for transendocardial cell delivery: the MyoStar^TM (Biosense Webster), the MyoCath™ (BioHeart), the Helix^TM (BioCardia), and the Stiletto^TM (Boston Scientific) catheter delivery systems (Fig. 14.2). The MyoStar™ catheter system utilizes a 3-dimensional (3-D) navigation system to guide injection, and the other three catheters are guided fluoroscopically therefore providing 2-dimensional orientation only. The first two catheter systems are introduced without a guidewire so require manipulation when advanced from the femoral artery and across the aortic valve while the second two catheters are introduced over a traditional guidewire system.

The MyoStar^TM mapping catheter system utilizes nonfluoroscopic guidance using the NOGA® (Biosense Webster) magnetic electromechanical system for intramyocardial navigation and mapping (Fig. 14.3). This technique uses ultralow-intensity magnetic fields generated by a triangular magnetic pad positioned beneath the patient and location sensor on the tip of the mapping catheter, which helps determine the real-time location and orientation of the catheter tip inside the left ventricle. The NOGA® system analyzes the movement of the catheter tip at the location of an endocardial point, timed with systole and diastole, and compares it with the movement of neighboring points – the percentage difference represents the degree of mechanical function at that endocardial point. The mapping catheter tip is also capable of measuring endocardial voltage potentials and an electrical map is constructed concurrently with the mechanical map. Healthy endocardium delivers a resting unipolar potential of 15 mV, whereas voltage values less than 6.9 mV reflect scar tissue; potentials between 7 and 15 mV suggest viable myocardium [72]. When the map is complete, all the data points are integrated by the NOGA® workstation and represented in a 3-D color-coded reconstruction of the endocardial surface which is able to distinguish between normal, ischemic (but viable), or infarcted myocardium (Fig. 14.3). Furthermore it allows accurate manipulation of the injection catheter in the left ventricle and therefore precise transendocardial injection. This allows targeted delivery of cells into specific areas, for example, BMSCs tend to be injected into the peri-infarct zone as it has been shown that BMSCs do not engraft well into scarred myocardium compared to skeletal myoblasts which seem capable of engrafting in fibrotic tissue [73]. The main disadvantages of this system are the additional time and costs required to produce an accurate electromechanical map to guide transendocardial injection together with additional operator training.

The MyoStar™ injection catheter (Fig. 14.4) incorporates an integrated injection needle and a location sensor at the tip which interfaces with the NOGA® 3-D mapping system. The 115-cm deflectable tip catheter consists of an outer shaft and inner core lumen, which runs the full length of the catheter and culminates distally in a nitinol 27-gauge needle. The core lumen can be advanced and retracted independently of other catheter movements. The proximal handle contains controls for catheter tip deflection, needle advancement, and the injection port for needle lumen injection. The catheter is prepared by adjusting the retractable needle extension length at 0° and 90° flex, setting the needle length to a maximal needle-to-wall ratio of 0.6. A 1-mL Luer Lock syringe containing the injectate is connected to the injection port, and the catheter “dead space” is filled with the cell suspension prior to introducing the catheter. Multiple injections can be performed at specific sites identified on the 3-D NOGA® electromechanical map – at our institution we perform circumferential 10 × 0.2-ml injections 1–2 cm apart into the hibernating myocardium around scar tissue. Needle injection into the true apex and myocardial segments with wall thickness less than 0.8 mm should be avoided to minimize the risk of perforation. This technology appears to be associated with a low complication rate in the preclinical and clinical studies performed [72, 74].

The MyoCath™ (BioHeart) injection catheter is similar to the MyoStar™ and consists of a 115-cm long deflectable injection catheter that also contains an integrated core lumen with a 25-gauge stainless steel needle at its distal end. This can also be advanced and retracted from the tip of the catheter and provides for multiple injections to a predetermined needle insertion depth. The MyoCath™ is however guided by fluoroscopic guidance, and additional imaging such as transesophageal echocardiography is sometimes utilized.

The Stiletto^TM (Boston Scientific) catheter (Fig. 14.2d) is also guided fluoroscopically and thus lacks the precision of the NOGA® system although real-time MRI guidance has shown promising results [75]. It contains three separate, independently moveable components: two steerable guiding catheters (9Fr and 7Fr) and an inner spring-loaded needle component. Catheter tip orientation is accomplished by guiding catheter manipulation, and the injection needle is set to a fixed depth (3.5 mm). The spring-loaded advancement mechanism theoretically allows the device to overcome fibrotic scar tissue resistance encountered during needle penetration. This technology is still under investigation; preclinical studies suggest that this is a “safe” technique [76] although higher than expected rates of pericardial effusion has led to the use of the Stiletto™ catheter being halted in the GENASIS investigational trial of VEGF-2 biologic for the treatment of refractory angina.

BioCardia’s Helix™ catheter (Fig. 14.5a) has a small, hollow, distal corkscrew needle, which is rotated into the endocardium to provide active fixation (similar to active fixation pacing leads) and to allow local delivery of biological therapeutic products. This catheter is used in combination with the BioCardia Morph® Deflectable (steerable) Guide Catheter (Fig. 14.5b) which helps catheter navigation within the LV under fluoroscopic guidance only. By eliminating the process of electromechanical mapping, this technique may have advantages in terms of procedural time and cost. Preclinical studies have proven safety and efficacy of this delivery method [77], and there is preliminary data to suggest that this method may be clinically effective [78].

Trans-coronary-venous intramyocardial (TVIM) injection can be performed using the Pioneer® catheter system (Medtronic Inc., now the Medtronic Pioneer catheter) that is placed percutaneously into the coronary sinus and target vein (Fig. 14.6). Using incorporated intravascular ultrasound (IVUS) guidance, a 24-gauge nitinol needle is extended to perform transvenous myocardial puncture. A 27-guage microinfusion catheter (IntraLume™) is then advanced though the needle to the targeted areas for cell delivery. Initial studies have confirmed the feasibility and safety of this approach in porcine models [79]. A phase 1 clinical trial has confirmed the safety and efficacy of this delivery method of skeletal myoblasts for the treatment of chronic ischemic LV systolic dysfunction [80]. Furthermore, an animal study demonstrated that the TVIM injection of BM-derived mononuclear cells may lead to a greater retention of the cellular product at the target area as compared to intracoronary infusion [81]. The main limitation to this approach is related to the technical difficulties associated with variation in coronary venous anatomy, which can make it difficult to selectively cannulate the target vein.

In summary, several techniques have been trialed for cell delivery in patients with acute and chronic ischemic heart disease. In animal models all routes appear to be equally effective in improving cardiac function although the exact mechanisms of action through which cell therapy may achieve this remains unanswered. There has been translation of all the delivery methods from animal models to humans, and to date the intramyocardial route appears to be the most promising in achieving higher levels of cell retention – whether this translates into improved clinical outcomes is yet to be seen.