Rexgenero Treats Its First Patient In Phase III Clinical Trial of Bone Marrow-Based Critical Limb Ischemia Treatment

According to their website, Rexgenero is a “regenerative medicine company and developer of advanced cell-based therapeutics for the treatment of serious diseases that are poorly treated with existing therapies.”   The head office of Rexgenero is in London, but their Research and Development laboratories are in Seville, Spain.  They also have a United Kingdom office in Brighton, which is a town with which I have some familiarity, since I lived there at one time in my life.

Rexgenero has focused on developing regenerative products from bone marrow cells.  In particular, they have used bone marrow-derived mononuclear cells (BM-MNCs).  BM-MNCs are extracted from the patient’s bone marrow after a bone marrow aspiration.  To isolate the mononuclear fraction, the white blood cells fraction from bone marrow is usually isolated by centrifugation with a material called Ficoll.  The pelleted material is the mononuclear fraction and contains a potent mixture of lymphocytes, monocytes, mesenchymal stem cells, hematopoietic progenitor cells, and other less well-characterized cell populations.


Another component of bone marrow mononuclear cells is endothelial progenitor cells (EPCs), which give rise to the walls of blood vessels.  When locally infused locally at the site of diseased vascular tissue, a well-processed mononuclear fraction can potentially form new blood vessels and restore tissue that was formerly starved for oxygen (see Haider, KH, Aziz S, and Al-Reshidi MA, Regenerative Medicine, 2017 12(8):969-982 for a review).

A bone marrow-based product developed by Rexgenero, called REX-001, has been used in human clinical trials to revascularize patients with a rather nasty condition called Critical Limb Ischemia (CLI).  CLI occurs when the blood vessels that feed and nourish an arm or leg are severely damaged or obstructed.  This results markedly reduces blood flow to the limb and the oxygen-deprived tissues can begin to die off, which causes excruciating pain, and disfiguring skin ulcers or sores.

CLI skin ulcer

These ulcers can become infected, gangrenous, and may require debridement or, in the most severe cases, limb amputation (yikes!).

Can REX-001 administration treat patients with CLI?  In an earlier Phase II clinical trial, which was completed in 2016, REX-001 was administered to over 100 patients in three clinical trials.  The results were rather encouraging. Patients who had severe ischemic pain at rest without skin ulcers (Rutherford scale in category 4), or resting pain with skin ulcers (category 5) were treated with REX-001 for 12 months.  At the end of 12 months, most patients (the website does not say what percentage of patients) in both patient populations were devoid of CLI and showed significant improvement in their clinical condition, as assessed by changes in Rutherford category.  Patients showed complete ulcer healing and alleviation of rest pain.  Angiographic imaging of treat patients definitively showed that REX-001-treated patients had extensive new networks that blood vessels that had sprouted.  It is reasonable, in my view to suspect that these new blood vessel networks in the limb were responsible for the corresponding improvement in patient’s clinical condition.

REX-001-induced new vasculature
Angiograms showing blood vessels prior to treatment (left) and after 6 months following treatment with REX-001 (right)

Rexgenero has recently announced the treatment of its first patient in a Phase III clinical trial that will evaluate REX-001, in patients with CLI and Diabetes Mellitus (DM).  In these two placebo-controlled, double-blind, adaptive Phase III trials, patients with CLI and DM with severe ischemic limb pain will be treated with REX-001 or a placebo to assess the efficacy and safety of REX-001 in the relief of CLI-associated resting limb pain.  In a second leg to this Phase III study, CLI patients with ischemic limb pain and skin ulcers (Rutherford stage 5) will be treated with REX-001 or a placebo to ascertain the efficacy and safety of REX-001 in the healing of CLI-induced skin ulcers. The relief of pain and the complete healing of ulcers are the primary endpoints of this clinical trial and amputation-free survival is the secondary endpoint in both legs of the study.

The European Medicines Agency (EMA) has fully endorsed the design of both legs of the trial, and this includes the target patient population and primary endpoints.  Clinical improvement in CLI patients increases the probability of successful treatment. Rexgenero also plans to enroll a total of 138 patients at approximately 35 clinical sites in Europe with first interim results expected in circa 18 months’ time and full data in 2020.

Joe Dupere, Rexgenero’s CEO, said “We are extremely pleased to announce the first patient infusion in our Phase III program with REX-001, which if successful could significantly improve the treatment of patients with CLI.”  He continued: “The program has been designed following advice from the EMA and in close collaboration with our Scientific Advisory Board. CLI is a medical condition with a clear need for new improved treatment options. Our goal is to bring innovative cell, gene and tissue therapies to the market addressing high unmet needs which cannot be treated with available therapies. We believe that REX-001 has the potential to be one of the first effective cell therapy products available for patients with CLI.”

This is definitely one clinical trial to keep an eye on.  If bone-marrow-based cell preparations can help patients sprout new blood vessel networks, there might be applications for other medical conditions as well, including cardiac ischemia, intestinal blood vessel blockage, and liver disease too.


Artificial blood vessels made by University of Minnesota Scientists

In patients who must receive dialysis to accommodate failing kidneys, ports are placed in their blood vessels, and a vein and an artery are tied together.  The name for the connection of an artery and a vein is a Cimino-Brescia fistula. Such fistulas are necessary for dialysis, and they are usually made in the arm. Since blood, like other fluids takes the path of least resistance, such fistulas generate high volume flow rates. Blood flow will prefer the fistula over capillary beds, which are high resistance flow areas. Also, native blood vessels are usually used to generate these fistulas because they are less likely to narrow and fail. Unfortunately, these surgical connections tend to fail. Worse still, they cannot be used in some patients because of the bad shape of their vascular system. Therefore, the answer in those cases is a graft. That seems onerous and likely to fail too.  Is there a better way?

Zeeshan H. Syedain and his coworkers from the laboratory of Robert Tranquillo at the University of Minnesota have used tissue engineering approach to generate vascular grafts from fibrin scaffolds and skin-based human fibroblasts.  In short, Tranquillo and his colleagues have made “off-the-shelf” blood vessels that were grown in the laboratory and do not have any living cells. Such lab-grown vessels might serve as blood vessel replacements for hard-up dialysis patients and others.  Tranquillo and his group published their findings in the journal Science Translational Medicine.

To make blood vessel substitutes, Tranquillo and others embedded human skin cells in a gel-like material made of cow fibrin. This concoction was grown in a bioreactor for seven weeks, after which, the cells were washed away. This left vessel-like tubes made of collagen and other proteins secreted by the cells.

Synthetic blood vessels
Researchers at the University of Minnesota have created a new lab-grown blood vessel replacement that is the first-of-its-kind nonsynthetic, decellularized graft that becomes repopulated with cells by the recipient’s own cells when implanted. Image courtesy of University of Minnesota.

Tranquillo said of this study, “We harnessed the body’s normal wound-healing system in this process by starting with skin cells in a fibrin gel, which is Nature’s starting point for healing.” He continued, “Washing away the cells in the final step reduces the chance of rejection. This also means the vessels can be stored and implanted when they are needed because they are no longer a living material.”

The vessel-like tubes looked like blood vessels, and they lacked any human cells.  Therefore, the immune system should not reject them if they were implanted into a human body.  However, can they function as blood vessels? To address this concern, Tranquillo and others implanted their laboratory-produced tubes into adult baboons. Six months after transplantation, the engrafted vessels looked like blood vessels and healthy cells from the recipient had grown into them and seemed to adapt to them without any ill effects. These laboratory-made vessels could withstand 30 times the average human blood pressure without bursting.  Additionally, there was no indication of an immune response and the grafts even self-healed when punctured with a needle.

Tranquillo and the team are in the process of FDA approval to test their synthetic blood vessels in clinical trials. In particular, Tranquillo and his team would like to test them in children with pediatric heart defects.

Infusion of high-dose umbilical cord blood cells normalized brain connectivity and improves motor function in children with cerebral palsy

Cerebral palsy is a congenital disorder that adversely affects movement, muscle tone and posture. Because those who suffer from congenital cerebral palsy are bone with it, there is often little that can be done to predict or prevent it. Cerebral palsy or CP is usually due to abnormal brain development prior to birth, but it can also result from in utero strokes, or oxygen deprivation during development or delivery. CP causes exaggerated reflexes, floppy or rigid limbs, and involuntary motions and there is a generalized weakness of skeletal muscles. CP affects 2-3 per 1,000 live births and the investment required by schools to accommodate CP children is substantial.  Furthermore, the personal investment of the heroic parents of CP children is substantial and, at times, exhausting.

Fortunately, animal models of CP have shown that the infusion of stem cells into the brains of young AP animals improves motor (movement-based) function.  In particular, human umbilical cord blood cells seem to facilitate repair of neural networks in the brain and improve movement. One study (Pediatric Research 2006; 59(2): 244-249) by Carola Meier and others from Ruhr-University in Bochum, Germany used an oxygen-deprivation model of CP in rats to show that treatment of these animals with human Umbilical Cord Blood Cells (hUBCs) substantially alleviated spastic paresis as assessed by walking track analysis. Also, examination of brain slices established that administered hUCBs incorporated themselves around the brain lesion (a phenomenon called “homing”) in large numbers. This study showed that the administration of hUCB stem cells after perinatal brain damage to could significantly reduce potential motor deficits.  A second paper (Developmental Neuroscience 2015;37(4-5):349-62) by Drobyshevsky and others from North Shore University Health System in Evanston, IL and collaborators from Duke University used a CP rabbit model to assess the efficacy of hUBCs to treat CP. In this experiment, Drobyshevsky and others induced oxygen deprivation when the rabbits were at 70% of their in utero lives. Then a group of the newborn rabbits were treated with hUCBs while others were not. The hUBC-treated animals showed significant improvements in posture, righting reflex, locomotion, tone, and dystonia (involuntary muscle contractions that cause repetitive or twisting movements). Unfortunately, the swimming test however showed that joint function was not restored by the hUBC treatment, but these other functions were. Tracking studies of the infused hUBC cells did not indicate that the cells penetrated into the brain with any efficiency, and Drobyshevsky and others suggested that the cells exerted their beneficial effects by means of “paracrine signaling,” which is to say that the cells secreted molecules that induced healing by activating native cells rather than differentiating into new neurons that created neural networks.


On the strength of these animal experiments, Jessica Sun from Duke University Medical Center and her colleagues and collaborators from multiple institutions extended these studies into human CP patients.  This, I’m sure, was a very dicey experiment to run because the subjects were children.  Getting approval for clinical experiments on children is very difficult and time-consuming.  Sun and her colleagues had shown that her hUBC infusion protocol was safe in a previous publication (Transfusion 2010; 50: 1980-1987) in which Sun and others reported treating 184 children with a single infusion of their own umbilical cord blood. The paper reported that the adverse effects of this treatment were rare and minimal. Because this was a Phase I study, it was only designed to assess the safety of the hUBC infusions and not their efficacy.


In a second publication, Sun and others have reported the results of their Phase II study in which they treated 63 CP children with various doses of hUBCs. This was a rigorous double-blind, placebo-controlled, crossover study in which Sun and her colleagues gave 10-50 million hUBCs per kilogram body weight to CP children between the ages 1 to 6 years.  These children received either their own umbilical cord blood or a placebo at the start of the experiment, followed by an alternate infusion 1 year later. After 1 year, those children who had received their own UBCs at the beginning of the trial, received the placebo, and those who had received the placebo at the start of the trial received their own UBCs. The children were assessed by means of specific motor function tests and their brains were imaged by means of magnetic resonance imaging brain connectivity studies. These assessments were done at the start of the trial, and then 1, and 2 years after the treatment.  To assess their motor skills, children were tested with a clinical tool called the Gross Motor Function Measure-66 tool.  This clinical tool evaluates changes in motor function in CP children.  Children are asked to perform a range of everyday activities from lying and rolling to walking, running, and jumping.  The children are given a composite score for all 66 tasks they are asked to do and this score reflects the depth of their motor skill. Changes in the Gross Motor Function Measure-66 (GMFM-66) indicates an improvement or decrease in motor function.  The primary endpoint was change in motor function 1 year after baseline infusion.

Two years after the initial treatments, the children were given further evaluations. Of the 63 CP children, 32 received their own umbilical cord blood and 31 received the placebo at the start of the experiments. One year after the trial began, Sun and her team detected no average change in GMFM-66) scores between the placebo and treated groups.  However, two years after the start of the trials, those CP children who had received higher doses of their own umbilical cord blood (20 million cells or more) showed significantly greater increases in their GMFM-66 scores.  In fact, their GMFM-66 scores were above what CP children at this specific age usually score. Another test that was administered was the Peabody Developmental Motor Scales test, which consists of six subtests that measure abilities in early motor development and assesses gross and fine motor skills in children from birth through five years of age. Gross Motor Quotient scores from the Peabody Developmental Motor Scales tests also revealed that children who had received the higher dose UBC treatments showed normalized scores, which indicates that the motor development of these children had become more normal rather than delayed.

Finally, the MRIs revealed normalized brain connectivity in the CP children who had received the higher doses of their own umbilical cord blood cells.

While this study is still preliminary, it suggests that appropriately doses of a child’s own umbilical cord blood stem cells improves brain connectivity and gross motor function in young children with CP.


Making Blood Cells in Culture – Done


One of the “Holy Grails” of stem cell biology has been growing blood cells in culture for use in clinical settings. Such a feat would provide large quantities of blood cells for post-surgical patients, or those with leukemia or other blood-based illness. The clinical applications are manifold and extensive.

Unfortunately, growing blood-making stem cells in the laboratory has proven to be a difficult task for even the most inventive and skilled stem cell laboratories. Nevertheless, several laboratories have been able to recapitulate the differentiation of pluripotent stem cells into cells that have the capacity to form T-cells and myeloid (non-lymphoid) cells (see Kennedy, M. et al. Cell Rep. 2, 1722–1735 (2012); Ditadi, A. et al. Nat. Cell Biol. 17, 580–591 (2015); and Elcheva, I. et al. Nat. Commun. 5, 4372 (2014)). Unfortunately, these experiments generated cells that were not able to engraft in the bone marrow of irradiated mice. Such an experiment is essential because radiation destroys the bone marrow of the mouse, and if a cultured cell is indeed and blood-cell-forming stem cell, then placing it into the bone marrow of irradiated mice should result in a functional restoration of the bone marrow. This, however, was not the case, which shows that whatever these pluripotent stem cells in these experiments differentiated into, they were not blood-cell-forming hematopoietic stem cells (HSCs).

Now, after a hiatus of almost 20 years, two different research groups have use two very different approaches to transform mature cells into primitive HSCs that are self-propagating and also form the cellular components of blood.

The first of these research teams was led by George Daley of Boston Children’s Hospital in Massachusetts. Daley’s group used induced pluripotent stem cell technology to reprogram adult human cells into cells that function as HSCs, even though they are not precisely like those found in the bone marrow in people. The second research team was led by Shahin Rafii of the Weill Cornell Medical College in New York City. Rafii and his coworkers used direct programming to differentiate mature cells from mice into fully functional HSCs.


The Daley group isolated skin-based fibroblasts from adult donors and then reprogrammed then through a combination of genetic engineering and cell culture techniques. This technology is similar to that designed by Shinya Yamanaka and his colleagues at Kyoto University, for which Yamanaka won the Nobel Prize in Medicine in 2012.   Once these reprogrammed cells formed induced pluripotent stem cells (iPSCs), Daley and his group did something very creative. They inserted the genes that encode seven different transcription factors into the genomes of their iPSCs. Transcription factors are proteins that activate gene expression. Transcription factors do so either by binding specific sequences of DNA, or by tightly binding to other proteins involved in gene expression and activating them. The genes for these seven transcription factors (ERG, HOXA5, HOXA9, HOXA10, LCOR, RUNX1 and SPI1) are known to be sufficient to convert hemogenic endothelium (the cells from which HSCs develop) into HSCs.

After engineering their iPSCs with these seven genes, Daley’s group did yet another highly creative thing. Daley and his colleagues injected their modified human cells into developing mice. This provided the cells with the proper environment to differentiate into HSCs. Twelve weeks after injecting them into mouse embryos, the engineered iPSCs had differentiated into progenitor cells that could produce the full range of blood cells found in human blood. This included immune cells, platelets, and other types of red and white blood cells. These progenitor cells are, according to Daley, “tantalizingly close” to naturally occurring HSCs.

Rafii and his team made their mouse HSCs from mature mouse cells without going through an embryonic intermediate. The Rafii grouop isolated endothelial cells that line blood vessels from adult mice and genetically engineered them to overexpress four different genes (Fosb, Gfi1, Runx1, and Spi1). Upon culturing their genetically engineered cells in a culture system that mimicked blood vessels, these cells, over time, differentiated into HSCs.

For the next test, Rafii and others injected their cultures-derived HSCs into irradiated mice. These mice survived and showed a completely recapitulated bone marrow that produced immune cells, and all types of red and white blood cells, and lived than 1.5 years in the lab.

Rafii told Nature’s Amy Maxmen that his approach is like “a direct airplane flight, and Daley’s procedure to a flight that takes a detour to the Moon before reaching its final destination.” Rafii noted that how the cells are made matters when it comes to using them in the clinic. Every time genes are transformed into cultured cells, a significant percentage of the cells fail to incorporate one or all of these genes. Such cells must be removed from those cells that were successfully transformed. Genetically engineered cells also run the risk of having experienced mutations as a side effect of the genetic manipulation. If implanted into people, such cells might cause problems.

Daley, however, and other stem cell researchers remain sanguine about the possibility of making such cells in safer, more efficient and even cheaper ways that can be brought to the clinic. For example, Jeanne Loring from the Scripps Research Institute in La Jolla, CA has suggested that using techniques that cause transient rather than permanent expression of introduced genes might very well make such cells inherently safer. Loring also noted that the iPSCs had by Daley’s group are initially made from skin-based fibroblasts, which are easy to acquire and isolate, whereas Rafii’s method begins with endothelial cells, which are more difficult to gather and to keep alive in the lab.

“For many years, people have figured out parts of this recipe, but they’ve never quite gotten there,” says Mick Bhatia, a stem-cell researcher at McMaster University in Hamilton, Canada, who was not involved with either study. “This is the first time researchers have checked all the boxes and made blood stem cells.” Bhatia added: “A lot of people have become jaded, saying that these cells don’t exist in nature and you can’t just push them into becoming anything else. . . I hoped the critics were wrong, and now I know they were.”

So making blood cells and HSCs in the laboratory is possible.  Bring this into the clinic is going to be even tougher.

Repeated Administrations of Stem Cells are More Effective than Single Administration

After a heart attack, the heart can undergo several structural and functional changes. Even after oxygen delivery to the heart muscle has been restored, a temporary loss of contractile function can persist for several hours or even days. This phenomenon is called myocardial stunning and it can also occur in people who have undergone cardiovascular procedures or central nervous system trauma. Myocardial stunning seems to result from the release of toxic molecules by the dying heart muscle cells, and an imbalance in ions required for heart muscle contraction, such as calcium ions.

If the heart muscle remains deprived of oxygen for some time, then the heart muscle adapts to low-oxygen conditions and “hibernates.” Hibernating myocardium contracts very little, and has “battened down the hatches,” metabolically speaking, in order to survive. However, hibernating myocardium takes even more heart muscle cells off-line and further reduces the performance of the heart.

Reduced heart performance leads to the production of molecules by the blood-starved kidneys that enlarge the heart and further compromise its efficiency. This leads to congestive heart failure and death. The term for such post-heart attack heart disease is ischemic cardiomyopathy” and it refers to a heart after a heart attack that is deprived of oxygen in many places and has heart walls filled with dead tissue that struggles to properly supply to body with blood, and often deteriorates as a result of this struggle.

Several different cell types have been used in many different studies to treat ischemic (oxygen-deprived) heart disease. The results, though positive in many cases, only show modest improvements in most cases. Furthermore, clinical studies and pro-clinical studies in laboratory animals tend to produce inconsistent results in which some patients or animals significantly improve while others either fail to improve at all. However, all of these studies have one feature in common: the subject is treated with only one infusion of stem cells. What if only one treatment is not enough to properly heal the damaged heart after a heart attack?

Workers from Roberto Bolli’s laboratory at the University of Louisville, Kentucky have treated laboratory rodents with heart attacks with three doses of cardiac progenitor cells. These treatments were given 35 days apart and were compared with single treatments and placebo treatments.

In their paper, which was published in Circulation Research (DOI:10.1161/CIRCRESAHA.116.308937), Bolli and his group gave heart attack to 85 Fischer 344 rats, but 13 died one week after the procedure. The remaining 72 rats were split into three groups: vehicle only, single-treatment, and triple-treatment. The vehicle-only rats received injections of saline into their heart muscle 30 days after the heart attack, and then again 35 days later and then again 35 days after that. The single-treatment rats received an injection of 12 million cardiac progenitor cells into their heart muscle 30 days after their heart attacks, but then received injections of saline 35 days later and 35 days after that. The triple-treatment rats received an injection of 12 million cardiac progenitor cells into their heart muscle 30 days after their heart attack, and then another injection 35 days later and a third inject 35 days after that. Of the 72 rats that survived the laboratory-induced heart attacks, 9 died as a result of injections into the heart. Thus, 63 rats completed the protocol.

Cardiac progenitor cells (CPCs) are resident stem cells in the heart that can be isolated with small heart tissue biopsies (see Tang XL, et al., Circ Res 2016;118:1091-1105). They can differentiate into heart muscle, blood vessels, or other heart-specific cell types (See Parmacek and Epstein, Cell 2005;120;295-298). Bolli and his co-workers have shown that the infusion of CPCs into the hearts of laboratory rodents after a heart attack can improve heart function, but the CPCs do not engraft into the heart at a terribly high rate. Furthermore, the functional improvements in the heart of these rodent elicited by CPC implantation are long-term (see Circ Res 2016;118:1091-1105).

When heart function of the laboratory rats were assessed prior to the procedure, no significant differences were observed between any of the rats in the three groups. However, after the procedure, the hearts of those rats that were injected with saline continued to deteriorate throughout the duration of the experiment. This deterioration was functional in nature and structural.

Animals that had received only one injection of 12 million CPCs into the left ventricles of their hearts showed significant improvements over the saline-injected animals. These animals showed less dilation of the heart (lower end-systolic volume), their hearts pumped more blood per beat (stroke volume), better thickness in the damaged heart wall, and improved ejection fraction (average percentage of blood pumped from the left ventricle during each beat). These hearts also had smaller heart scars, great elasticity, and greater amounts of viable muscle.

While all that sounds great, the triply-injected hearts that received three injections of CPCs, showed even more robust and significant improvements in heart structure and function. Whatever the single injection of CPCs did, the triple injections did even better. However, it must be noted that even with three injections of CPCs, the level of engraftment of the injected cells remained poor.

To summarize these results observed in this paper, the 3 doses of CPCs 35 days apart resulted in increases in local and global heart function that were, roughly, triple that produced by a single dose. The multiple CPC administrations were associated with more viable tissue, less scar tissue, less collagen, and greater heart muscle cell density in the infarcted region. Still, the level of engraftment and differentiation of the injected cells accounted for <1% of total heart muscle cells.

Bolli and his coworkers believe that their work suggests that all clinical and preclinical trials should at least try multiple stem cell treatments in order to maximize the clinical benefit of the injected stem cells. Furthermore, Bolli and others suggest that the use of single injection protocols are the reason so many stem cell-based clinical trials have resulted in inconsistent and inconclusive results.

This study is a large preclinical trial that used large numbers of animals. The data are solid and the results are believable. My problem with the clinical implications of this study are as follows: A heart attack causes a good deal of inflammation in the heart that culminates in wound healing. Within 24 hours of the heart attack, white blood cells infiltrate the damaged area of the heart. Protein-degrading enzymes from scavenger neutrophils (a type of white blood cell) degrade dead tissue. The damaged cells degenerate, and collagen-making fibroblasts divide and lay down scar tissue. The initially-deposited collagen deposited in the wall of the heart is weak, mushy, and vulnerable to re-injury. Unfortunately, this is precisely the period of time (10-14 days after the heart attack) that patients feel better and want to increase their activity levels. However, this greater activity level can stress the heart and cause rupture of the heart wall. After 6 weeks, the dead (necrotic) area is completely replaced by scar tissue, which is strong but incapable of contracting or relaxing.

In light of this timeline, multiple injections of stem cells into the heart after a heart attack might very increase the risk of rupture of the heart wall. This is particularly the case if implanted cells secrete tissue proteases that degrade surrounding tissue. Thus, timing and dose will be extremely important in such multiple treatments. Too many cells can rupture the heart, treatments too early when the healing heart walls are sift and weak will prove inimical to the heart and treatments given too late might very well be too late for the cells to do any good. Therefore, while this paper seems to move the ball down the field of regenerative medicine, it creates a fair number of questions that will need to be answered before such a strategy can come to the clinic.

Dosing Recent Heart Attack Patients with G-CSF Doesn’t Seem To Work

Granulocyte-Colony Stimulating Factor (G-CSF)is a small protein that stimulates the bone marrow to produce more of a particular class of white blood cells called granulocytes and release them into the bloodstream. A commercially available version of G-CSF called Filgrastim (Neupogen) is used to boost the immune system of cancer patients whose immune systems have taken a beating from chemotherapy.

Because several clinical trials have shown that implanting bone marrow mononuclear fractions into the hearts of heart attack patients can improve the heart health of some heart attack patients, clinicians have supposed that injecting heart attack patients with drugs like filgrastim, which moves many bone marrow-derived cells into the bloodstream might also provide some relief for heart attack patients.

Nice idea, but it does not seem to work. Two clinical trials, STEMMI and REVIVAL-2, have given G-CSF to heart attack patients at different times after their heart attacks. Unfortunately both studies have failed to show a difference from the placebo.

In the REVIVAL-2 study, 114 patients were enrolled, and 56 received 10 micrograms per kilogram body weight G-CSF for five days, and the remaining patients received a placebo treatment.  G-CSF and the placebo were administered to patients five days after the hearts were successfully reperfused by percutaneous coronary intervention (this is a fancy way of saying stenting).  This study was double-blinded, placebo-controlled and well designed.  Unfortunately, when patients were studied seven years after treatment, there were no statistically significant differences between the treatment and the placebo groups when it came to the number of deaths, heart attacks, and strokes.  Thus, the authors conclude that G-CSF administration did not improve clinical outcomes for patients who had a heart attack (see Birgit Steppich, et al, Atherosclerosis and Ischemic Disease 115.4, 2016).

A second clinical trial, the STEMMI trial, was a prospective trial in which G-CSF treatment was begun 10-65 hours after reperfusion.  Here again, there were no structural differences between the placebo group and the G-CSF-treated group six months after treatment and a five-year follow-up analysis of 74 patients revealed no differences in the occurrence of major cardiovascular incidents between the two treatment groups (R.S. Ripa, and others, Circulation 2006; 113: 1983-1992).

The STEM-AMI clinical trial also showed no differences in clinical outcomes after G-CSF treatment as compared to placebo in 60 patients after three years (F. Achilli, and others, Heart 2014, 100: 574-581).

Why does this technique fail?  It is possible that the white blood cells that are mobilized by G-CSF are low-quality and do not express particular genes.  A study in rats has shown that G-CSF infusion increases the number of progenitor cells in the bloodstream, but fails to increase the number of progenitor cells in the heart after a heart attack (D. Sato, and others, Experimental Clinical Cardiology, 2012; 17:83-88).  In order for cells to home to the infarcted heart, they must express particular proteins on their surfaces.  For example, the cell surface protein CXCR4 is known to play an integral role in progenitor cell homing, along with several other proteins (see Taghavi and George, American Journal of Translational Research 2013; 5:404-411; Shah and Shalia, Stem Cells International 2011;2011:536758; Zaruba and Franz, Expert Opinion in Biological Therapy 2010; 10:321-335).  Indeed, Stein and others have shown that progenitor cells mobilized with G-CSF in human patients lack CXCR4 and other cell adhesion proteins thought to play a role in homing to the infarcted heart (Thromb Haemost 2010;103:638-643).

Therefore, even though all of these studies have not uncovered a risk in G-CSF treatment, the consensus of the data seems to be there no clinical benefit is conferred by treating heart attack patients with G-CSF.