Three New Clinical Trials Examine Bone Marrow-Based Stem Cells To Treat Heart Failure

In April of 2013, the results of three clinical trials that examined the effects of bone marrow-derived stem cell treatments in patients with acute myocardial infarction (translation – a recent heart attack) or chronic heart failure. These trials were the SWISS-AMI trial, the CELLWAVE trial, and the C-CURE trial.

The SWISS-AMI trial (Circulation. 2013;127:1968-1979), which stands for the Swiss Multicenter Intracoronary Stem Cells Study in Acute Myocardial Infarction trial, was designed to examine the optimal time of stem cell administration at 2 different time points: early or 5 to 7 days versus late or 3 to 4 weeks after a heart attack. This trial is an extension of the large REPAIR-AMI, which showed that patients who tended to receive bone marrow stem cell treatments later rather than earlier had more pronounced therapeutic effects from the stem cell treatments.

SWISS-AMI examined 60 patients who received standard cardiological care after a heart attack, 58 who received bone marrow stem cells 5-7 days after a heart attack, and 49 patients who received bone marrow stem cells 3-4 weeks after their heart attacks. All stem cells were delivered through the coronary arteries by means of the same technology used to deliver a stent.

When the heart function of all three groups were analyzed, no significant differences between the three groups were observed. Those who received stem cell 5-7 days after a heart attack showed a 1.8% increase in their ejection fractions (the percentage of blood that is ejected from the ventricle with each beat) versus an average decrease of 0.4% in those who received standard care, and a 0.8% increase in those who received their stem cells 3-4 weeks after a heart attack. If these results sound underwhelming it is because they are. The standard deviations of each group so massive that these three groups essentially overlap each other. The differences are not significant from a statistical perspective. Thus the results of this study were definitely negative.

The second study, CELLWAVE (JAMA, April 17, 2013—Vol 309, No. 15, 1622-1631), was a double-blinded, placebo-controlled study conducted among heart attack patients between 2005 and 2011 at Goethe University Frankfurt, Germany. In this study, the damaged area of heart was pretreated with low-energy ultrasound shock waves, after which patients in each group were treated with either low dose stem cells, high-dose stem cells, or placebo. Patients also received either shock wave treatment or placebo shock wave treatment. Thus this was a very well-controlled study. Stem cells were administered through the coronary arteries, just as in the case of the SWISS-AMI study.

The results were clearly positive in this study. The stem cell + shock wave treatment groups showed definite increases in heart function above the placebo groups, and showed fewer adverse effects. The shock wave treatments seem to prime the heart tissue to receive the stem cells. The shock waves induce the release of cardiac stromal-derived factor-1, which is a potent chemoattractor of stem cells.  This is an intriguing procedure that deserves more study.

The third study, C-CURE, is definitely the most interesting of the three (Bartunek et al. JACC Vol. 61, No. 23, June 11, 2013:2329–38). In this trial, mesenchymal stromal cells (MSCs) were isolated from bone marrow and primed with a cocktail of chemicals that pushed the stem cells towards a heart muscle fate. Then the cells were transplanted into the heart by direct injection into the heart muscle as guided by NOGA three-dimensional imaging of the heart.

After initially screening 320 patients with chronic heart failure, 15 were treated with standard care and the other 32 received the stem cell treatment. After a two-year follow-up, the results were remarkable: those who received the stem cell treatment showed an average 7% increase in ejection fraction versus 0.2% for receiving standard care, an almost 25 milliliter reduction in end systolic volume (measures degree of dilation of ventricle – not a good thing and the fact that it decreased is a very good thing) versus a 9 milliliter decrease for those receiving standard care, and were able to walk 62 meters further in 6 minutes as opposed to standard care group who walked 18 meters less in 6 minutes.

While these studies do not provide definitive answers to the bone marrow/heart treatment debate, they do extend the debate. Clearly bone marrow stem cells help some patients and do not help others. The difference between these two groups of patients continues to elude researchers. Also, how the bone marrow is processed is definitely important. When the cells are administered also seems to be important, but the exact time slot is not clear in human patients. It is also possible that some patients have poor quality bone marrow in the first place, and might be better served by allogeneic (someone else’s stem cells) treatments rather than autologous (the patient’s own stem cells) stem cell treatments.

Also, stem cell treatments for heart patients will probably need to be more sophisticated if they are to provide greater levels of healing. Heart muscle cells are required, but so are blood vessels to feed the new heart muscle. If mesenchymal stem cells work by activating resident heart stem cells, then maybe mesenchymal transplants should be accompanied by endothelial progenitor cell transplants (CD117+, CD45+ CD31+ cells from bone marrow) to provide the blood vessels necessary to replace the clogged blood vessels and the new heart muscle that is grown.

For Treating Heart Attacks, Satellite Cells Lacking MyoD are Superior to Those With MyoD

Atsushi Asakura and his colleagues at the University of Minnesota Stem Cell Institute have extended some of their earlier findings in a paper that appeared in PLoS One last year. This paper is almost a year old by now, but its results are fascinating and are definitely worth examining.

In 2007, Asakura published a paper with the Canadian researcher Michael A. Rudnicki in the Proceedings of the National Academy of Sciences. In this paper, Asakura and his colleagues examined the ability of muscle satellite cells from MyoD- mice to integrate into injured muscle. I realize that last sentence just sounded like gobbledygook, to some of you, but I will try to put the cookies on a lower shelf.

Satellite cells constitute a stem cell population within skeletal muscle. They are a small population of muscle-making stem cells found in skeletal muscle and they express a whole host of muscle-specific genes (e.g., desmin, Pax7, MyoD, Myf5, and M-cadherin). Satellite cells are responsible for muscle repair, but previous work has shown that there are at least two populations of satellite cells in skeletal muscle. One population rapidly contributes to muscle repair, whereas the other population is more stem cell-like and remains longer in an undifferentiated state in the recipient muscle (see Beauchamp JR , et al (1999) J Cell Biol 144:1113–1122; Kuang S , et al (2007) Cell 129:999–1010). Presently, it is not clear which population is more efficient in repairing continuously degenerating muscle.

MyoD is a gene that encodes a protein that binds to DNA and activates the expression of particular genes. It plays a vital role in regulating muscle differentiation, and belongs to a family of proteins known as myogenic regulatory factors or MRFs. All MRFs are bHLH or basic helix loop helix transcription factors, and they act sequentially in muscle differentiation. MRF family members include MyoD, Myf5, myogenin, and MRF4 (Myf6). MyoD is one of the earliest genes that indicates a cell has committed to become a muscle cell. MyoD is expressed in activated satellite cells, but not in quiescent (sleeping) satellite cells. Strangely, even though MyoD marks myoblast commitment, muscle development is not dramatically prevented in mouse mutants that lack the MyoD gene. However, this is likely to result from functional redundancy from Myf5. Nevertheless, the combination of MyoD and Myf5 is vital to the success of muscle production.


Therefore, Asakura and his crew decided to isolated muscle satellite cells from mice that lacked functional copies of the MyoD gene. Making such mice is labor intensive, but doable with mouse embryonic stem cell technology. When such MyoD- mice were made, Asakura and others isolated the satellite cells from these mice and characterized them. They discovered in their 2007 paper, that the satellite cells from the MyoD- mice were much more stem cell-like than satellite cells from MyoD+ mice. The MyoD- satellite cells grew better in culture, integrated into injured muscles better and survived better than their MyoD+ counterparts.

Why is this important? Because when it comes to treating degenerative muscle diseases like muscular dystrophy, finding the best cell is crucial. MyoD+ satellite cells have been used, but they are limited in the amount of muscle repair they provide. MyoD- cells might be a better option for treating a disease like muscular dystrophy.

Or for that matter, what about the heart? Finding the right cell to treat the heart after a heart attack has proven difficult. There are some things bone marrow cells do well, and other things they do not do well when it comes to regenerating the heart. Likewise, there are some things mesenchymal do well and other things they do not do well when placed in a damaged heart. Can MyoD- satellite cells do a better job than either of these types of stem cells?

That was the question addressed in the 2012 Nakamura paper that was published in PLoS One. Clinical trials that have treated heart attack patients with injections of MyoD+ satellite cells into the heart have shown that such treatments can improve heart function, but usually only transiently. They also prevent remodeling of the heart after a heart attack. However, two larger studies failed to produce significant improvements in heart function compared to the placebo, and patients who received the satellite cell transplants were also susceptible to very fast heart beats (tachycardia). Because of these downsides, the excitement for transplanting muscle satellite cells into the heart has waned.

So how did MyoD- satellite cells do? All the laboratory animals used in this experiment (BALB/c mice) were given heart attacks, and injected with either MyoD+ or MyoD- satellite cells. The hearts of animals injected with MyoD- satellite cells were compared with animals whose hearts were injected with MyoD+ satellite cells.

In culture, the MyoD- satellite cells grew better than the MyoD+ cells. When injected into the heart, the MyoD- cells integrated into the heart muscle and spread throughout the heart muscle much more robustly than the MyoD+ cells. The MyoD- cells were also much less susceptible to cell death and survived better than their MyoD+ counterparts.

(A) These panels show MI induced by left coronary artery ligation. Wild-type and MyoD−/− myoblasts were directly injected into the peri-infarct regions of LV. After 1 week, X-gal staining of whole heart indicated that more MyoD−/− myoblasts engrafted than wild-type myoblasts (arrows). Arrowheads indicate left coronary artery ligation points. X-gal staining of cross sections indicated that more MyoD−/− myoblasts than wild-type myoblasts engrafted in both injured and uninjured areas of the heart. Arrows indicate engrafted lacZ+ wild-type and MyoD−/− myoblasts. Scale bars = 1 mm.
(A) These panels show MI induced by left coronary artery ligation. Wild-type and MyoD−/− myoblasts were directly injected into the peri-infarct regions of LV. After 1 week, X-gal staining of whole heart indicated that more MyoD−/− myoblasts engrafted than wild-type myoblasts (arrows). Arrowheads indicate left coronary artery ligation points. X-gal staining of cross sections indicated that more MyoD−/− myoblasts than wild-type myoblasts engrafted in both injured and uninjured areas of the heart. Arrows indicate engrafted lacZ+ wild-type and MyoD−/− myoblasts. Scale bars = 1 mm.

Functionally speaking for the heart, animals that had received transplantations of MyoD- satellite cells had higher ejection fractions, small areas of dead heart tissue, lower end systolic and end diastolic volumes, and more normal echocardiograms. Even though MyoD- cells differentiated into skeletal muscle and not heart muscle (no surprise there), the MyoD- cells induced a very substantial quantity of new blood vessels to sprout in the scar area.

(A) Two weeks post-transplantation, immunofluorescence staining of heart cross sections showed that the progeny of lacZ+ wild-type and MyoD−/− myoblasts formed nestin+ multinucleated skeletal myotubes. Laminin (red) indicates cardiomyocytes and skeletal myotubes. Arrows indicate lacZ+ donor cell-derived nuclei in nestin+ myotubes. (B) Comparison of the relative numbers of lacZ+/nestin+ myotubes for wild-type and MyoD−/− myoblasts 2 weeks after injection (n = 3).
(A) Two weeks post-transplantation, immunofluorescence staining of heart cross sections showed that the progeny of lacZ+ wild-type and MyoD−/− myoblasts formed nestin+ multinucleated skeletal myotubes. Laminin (red) indicates cardiomyocytes and skeletal myotubes. Arrows indicate lacZ+ donor cell-derived nuclei in nestin+ myotubes. (B) Comparison of the relative numbers of lacZ+/nestin+ myotubes for wild-type and MyoD−/− myoblasts 2 weeks after injection (n = 3).

From these experiments, it seems that the MyoD- satellite cells are superior to the MyoD+ satellite cells for treating heart after a heart attack. These cells secrete a whole host of factors that aid the heart in healing and also structurally support the heart and prevent remodeling.

Might it be possible to use such cells in human trials? Asakura notes that engineering MyoD- satellite cells would be impractical for human clinical purposes, but it might be possible to downregulate MyoD expression with drugs (bromodeoxyuridine) or other reagents (RNAi or Id protein transformation).

This work shows that there is a better way to use muscle satellite cells for heart treatments. It simply requires you to remove MyoD function, and the cells will grow and spread throughout the heart better, and more robustly augment heart function and healing.

See NakamuraY, et al PLoS One 7(7) 2012:e41736.

Treating the Heart with Mesenchymal Stem Cells: Timing and Dosage

Stephen Worthley from the Cardiovascular Investigation Unit at the Royal Adelaide Hospital in Adelaide, Australia and his colleagues have conducted a timely experiment with rodents that examines the effects of dosage and timing on stem cell treatments in the heart after a heart attack.

Mesenchymal stem cells from bone marrow and other sources have been used to treat the heart of laboratory animals and humans after a heart attack. The optimal timing for such a treatment remains uncertain despite a respectable amount of work on this topic. Early intervention (one week) seems offer the best hope for preserving cardiac function, but the heart at this stage is highly inflamed and cell survival is poor. If treatment is delayed (2-3 weeks after the heart attack), the prospects for cell survival are better, but the heart at this time is undergoing remodeling and scar formation. Therefore, stem cell therapy at this time seems unlikely to work. Human clinical trials seem to suggest that mesenchymal stem cell treatment 2-3 weeks after a heart attack does no good (see Traverse JH, et al JAMA 2011;306:2110-9). The efficacy of the delivering mesenchymal stem cells to the heart at these different times has also not been compared.

If that degree of uncertainty is not enough, dosage is also a mystery. Rodent studies have used doses of one million cells, but studies have not established a linear relationship between efficacy and dose, and higher dosages seem to plateau in effectiveness (see Dixon JA, et al Circulation 2009;120(11 Suppl):S220-9). High doses might even be deleterious.

So what is the best time to administer after a heart attack, and how much should be administered? These are not trivial questions. Therefore a systematic study is required and laboratory animals such as rodents are required.

In this study, five groups of rats were given heart attacks by ligation of the left anterior descending artery, and two groups of rats received bone marrow-derived mesenchymal stem cells immediately after the heart attack. The first group received a low dose (one million cells) and the second group received twice as many cells. The three other groups received their treatments one week after the heart attack. The third group received the low dose of stem cells received the low dose of cells (one million cells), and the fourth group received the higher dose (two million cells). The fifth group received no such cell treatment.

All mesenchymal stem cells were conditioned before injection by growing them under low oxygen conditions. Such pretreatments increase the viability of the stem cells in the heart.

The results were interesting to say the least. when assayed four weeks after the heart attacks, the hearts of the control animals showed a left ventricular function that tanked. The ejection fraction fell to 1/3rd the original ejection fraction (~60% to ~20%) and stayed there. The early high dose animals showed the lowest decrease in ejection fraction (-8%). The early low dose group showed a greater decrease in ejection fraction. Clearly dose made a difference in the early-treated animals with a higher dose working better than a lower dose.

In the later-treated animals, dose made little difference and the recovery was better than the early low dose animals. when ejection fraction alone was considered. However, when other measures were considered, the picture becomes much more complex. End diastolic and end systolic volumes were all least increased in the early high dose animals, but all four groups show significantly lower increases than the controls. The mass of the heart, however, was highest in the late high-dose animals as was ventricular wall thickness.

When the movement of the heart walls were considered, the early-treated animals showed the best repair of those territories of the heart near the site of injection, but the later-treated animals showed better repair at a distance from the site of injection. The same held for blood vessel density: higher density in the injected area in the early-treated animals, and higher blood vessel density in those areas further from the site of injection in the later-treated animals.

The size of the heart scar clearly favored the early injected animals, which the lower amount of scarring in the early high dose animals. Finally when migration of the mesenchymal stem cells throughout the heart was determined by using green fluorescent protein-labeled mesenchymal stem cells, the later injected mesenchymal stem cells were much more numerous at remote locations from the site of injection, and the early treated animals only had mesenchymal stem cells at the site of injection and close to it.

These results show that the later doses of mesenchymal stem cells improve the myocardium further from the site of the infarction and the early treatment improve the myocardium at the site of the infraction. Cell dosage is important in the early treatments favoring a higher dose, but not nearly as important in the later treatments, where, if anything, the data favors a lower dose of cells.

Mesenchymal stem cells affect the heart muscle by secreting growth factors and other molecules that aids and abets healing and decreases inflammation. However, research on these cells pretty clearly shows that they modulate their secretions under different environmental conditions (see for example, Thangarajah H et al Stem Cells 2009;27:266-74). Therefore, the cells almost certainly secrete different molecules under these conditions.

In order to confirm these results, similar experiments in larger animals are warranted, since the rodent heart is a relatively poor model for the human heart as it beats much faster than human hearts.

See James Richardson, et al Journal of Cardiac Failure 2013;19(5):342-53.

Stem Cell Fixes for the Heart

Two recent papers have provided very good evidence that pluripotent stem cells can help heal a heart that has experienced a heart attack. One of these papers used induced pluripotent stem cells from rats, and the other used embryonic stem cells.

The first paper comes from the laboratory of Yoshiki Sawa, who is a professor in the Department of Surgery at the Osaka University Graduate School of Medicine in Osaka, Japan. In this paper, Sawa’s group made induced pluripotent stem cells (iPSCs) from mice and cultured them under conditions known to induce differentiation into heart muscle cells. Beating cells were detected and grown on gelatin-coated plates with Delbecco’s medium. When these cells were tested for gene expression, they made all the same genes as those found in a mouse heart.

To get the cells to form sheets of heart muscle cells, Sawa and his team plated his iPSCs on UpCell plates that are coated with a chemical that causes the cells to adhere to it at normal temperatures, but when the temperature is dropped, the cells detach from the plate. Sawa used another innovation with this culture system; he grew cell without any sugar. This caused all the non-heart cells to die off. The result was a sheet of heart muscle cells that contracted in unison.

Next, the Sawa team took induced heart attacks in a Japanese rat strain. 2 weeks after suffering the heart attack, the sheet of heart muscle cells were placed on the heart scar in half of the rats and the other half received no implants.

Four weeks after implantation of the heart muscle sheet, the differences in heart function were stark. The ejection fraction in the hearts of the animals that had received the iPSC-derived heart muscle sheets increased almost 10%. The fractional shortening, which is the degree to which the heart muscle shortens when it contracts, also increased more than 5%. Also, the amount of stretching during pumping decreased, which indicates that the heart is pumping more efficiently.

When the heart muscle from the implants were examined, they were also filled with molecules associated with the production of new blood vessels. Thus the implanted heart muscle sheets also helped heal the heart by inducing the formation of new blood vessels.

A danger of using iPSC-derived heart muscle cells is the tendency to miss undifferentiated cells and have undifferentiated cells that cause tumors. In this experiment, they noticed tumors if they only grew the cells in the sugar-free medium for a little while. However, if they grew the iPSC-heart muscle cells in sugar-free media for at least three days, all the tumor-causing cells died and implants from these sheets never formed any tumors.

This paper demonstrated the efficacy and plausibility of using patient-specific iPSCs to treat a heart that has had a heart attack some time ago.

The second paper comes from the laboratory of Marisa Jaconi in Geneva, Switzerland. In this paper, Jaconi and her gang of stem cell scientists at the Geneva University Hospitals and the Ecole Polytechnique Fédérale de Lausanne used a “cardiopatch” seeded with cardiac-committed embryonic stem cells to treat a heart attack in rats.

Because the injection of stem cells can induce arrhythmias (irregular heart beats), narrowing of blood vessels, blood vessel obstruction, and other types of damage, these two papers tried to use sheets of cells or cells embedded in biodegradable patches to treat the heart. In this paper, Jacobi and others used a hydrogel made from fibrin, which is the same material found in blood clots. Into that fibrin hydrogel, they placed mouse embryonic stem cells that had been treated with a protein called BMP-2, which drives pluripotent stem cells toward a heart cell fate.

To use these cardiopatches, Jacobi and her group induced heart attacks in a French rat strain and then applied the patch to the heart. They had two groups of rats; those that had been given heart attacks and those that had not. The sham group received either a patch with cells, a patch with iron particles (for detection with MRI) or not patch. The heart attack group received the same.

The results are a little hard to interpret, but the patch + cells definitely improved heart function. First, the hearts that had received patches with cells showed in increase in small blood vessels and blood vessel-making (CD31+) cells. Therefore the patches + cells improved heart circulation. Second, the hearts with the patch + cells showed the presence of new heart muscle cells and much mess thinning of the walls of the heart. Third, the heart functional parameters were better preserved in the patch + cells hearts. The ejection fraction decreased substantially in the hearts that did not receive cells, but in the hearts that received patch + cells, the amount of blood left in the heart after pumping and at rest did not increase nearly as much as in the other groups. These parameters are in indication of the efficiency with which the heart is pumping. The fact that the heart + cells hearts did not decrease in efficiency nearly as precipitously as the others shows that the stem cells are healing the heart.

While these results may not seem terribly robust, we must remember that the cardiopatch was only placed over a small portion of the heart. Therefore, we would not expect to see large increased in function. The fact that we do see new heart muscle cells, new blood vessels, and an arrest in the functional free fall of the heart is significant, given the small area of the heart that was cover with the cells.

The cardiopatch is a new technology and this experiment showed that the patch biodegrades quickly and without incident. It also showed that embedding cells in the patch is feasible, and that the patch is a plausible vehicle to deliver cells to the heart. This procedure also induced the formation of new heart muscle cells in the heart scar and new blood vessels too. Perhaps even more encouraging is the absence of tumors reported in this paper. Even though the ESCs were not differentiated completely into heart muscle cells, the cardiac-directed cells were differentiated enough to form either blood vessels, smooth muscle, or heart muscle. This seems to be enough to prevent the cells from forming tumors. Also, the fibrin scaffold was not deleterious to the heart, even though some studies have used other scaffolds that are damaging to the heart.

Thus cardiopatches and cardiac muscle sheets are perfectly good strategies for treating heart with stem cells. More work needs to be done, but the results are encouraging.