Stephen E. Epstein, M.D.

After a string of failures, researchers are coming back from the drawing board with new strategies.

Beginning in about 1990, the treatment of cardiovascular disease experienced major conceptual expansions through the application of new therapeutic paradigms targeted to angiogenesis and tissue healing. Initial strategies involved administering proteins involved in healing processes, then the genes that encoded them, and then, by the end of the decade, the use of stem cells to better achieve these objectives.

Some of the first experiments showed that target tissues needed more prolonged exposure than single angiogenic proteins could provide. Subsequent experiments tested the effects of the genes that encoded these proteins, leading to prolonged expression of the encoded proteins. Many preclinical and preliminary clinical studies reported encouraging results, but the definitive, randomized clinical studies with humans were disappointing.

Early Cell Therapy Strategies

The clinical trial failures with protein and gene therapy led to more basic studies to uncover underlying mechanisms involved in healing and angiogenesis processes. In a mouse study (CW Lee et al., JACC 2004), we discovered that over 700 hundred genes were differentially expressed in the ischemic hindlimb over the two weeks following ligation of the femoral artery—a time during which collateral development and healing occurred. It’s likely not all of these gene products were involved in the healing response; however, what’s clear is that potentially hundreds of differentially expressed genes are involved. Researchers therefore questioned the ’90s strategy of administering a single protein or gene to enhance healing. This insight led to a focus on the use of stem cells since these cells produce multiple proteins involved in healing and angiogenesis.

From the ’00s until about 2012, stem cell therapy-based studies typically used autologous, or patient-derived, cells. An excellent meta-analysis (Clifford DM et al., Stem cell treatment for acute myocardial infarction. Cochrane Database of Systematic Reviewers 2012, Issue 2) evaluated the results of 39 randomized, controlled studies of acute myocardial infarction (AMI) cell-based therapy trials performed through 2012 and including over 1,700 patients. They found cell-based therapy significantly improved the heart’s pumping capacity by only 1.78 percent, an increase with questionable clinical relevance. Moreover, the meta-analysis revealed no improvement in mortality or morbidity. Similar disappointing results were found in randomized cell-based therapy trials involving patients with congestive heart failure (Sanganalmath and Bolli. Circ Res. 2013). These findings led to several important new directions many cell-based investigators in the cardiology field are taking.

The Use of Allogeneic MSCs

Past studies have shown that autologous cells obtained from older patients, patients with multiple risk factors, or both, have impaired function. Consequently, an increasing number of studies are using allogeneic cells, obtained from young healthy donors. In contrast to the host-based immune responses elicited by most cell types derived from an allogeneic source, mesenchymal stem cells (MSCs) are relatively immunoprivileged. Like most stem cells, MSCs are cytokine factories, secreting numerous proteins known to be involved in healing and angiogenesis processes. When these adult stem cells are used therapeutically, it is now generally recognized that any resulting improvement in cardiac function cannot be ascribable to differentiation of the stem cells into cardiac myocytes. Rather, any benefit almost certainly results from the paracrine effects stem cell factors have on injured tissue. (Kinnaird, Epstein et al., Circ 2004)

Aging and Risk Factor-Induced Compromise of Stem Cell Functionality

Stem cells have the capacity to secrete a broad array of factors with important physiologic effects. This function is impaired by aging and in individuals with various risk factors for coronary artery disease (Kinnaird, Burnett, Epstein et al., Cardiovascular Research 2008). One example is the aging-induced impairment of the hypoxia-inducible-factor (HIF) signaling pathway.

This intracellular pathway is activated by hypoxia, which stimulates expression of genes designed to combat the effects of hypoxia. Thus, in the presence of hypoxia a heterodimer—consisting of HIF-1-beta and HIF-1-alpha—is formed that, along with p300, leads to the transcription of multiple genes critical to the healing responses triggered by ischemia-induced tissue injury.

We observed that the normal increase in HIF expression occurring in MSCs in response to hypoxia was severely impaired in MSCs obtained from old versus young mice (Kinnaird, Burnett, Epstein et al., Cardiovascular Research 2008). The Western blot on the left side of Figure 1 shows HIF-1-alpha levels in MSCs obtained from young and from old mice under normoxic conditions. As expected, HIF-1-alpha levels are low. Also as expected, when MSCs derived from young mice are exposed to hypoxia, HIF-1-alpha levels increase markedly. However, HIF-1-alpha expression was markedly impaired in MSCs obtained from old mice.

Figure 1. These Western blots show that the MSCs of old mice—when exposed to 12 hours of hypoxia – express significantly less HIF-1 compared with young mice. This results in impaired expression of genes that are downstream to HIF-1, including multiple genes relating to angiogenesis and tissue healing.

The Value of Culturing MSCs under Hypoxic Conditions

The recent concept of optimal oxygen levels for growing and expanding stem cells will probably have a major impact on adult stem cell therapeutics. Although arterial blood oxygen saturation is 100%, tissue consumption of oxygen leads to very low tissue oxygen concentrations. Hypoxia is, in a sense, normoxia for bone marrow cells because normally these cells live in an oxygen concentration of 1–6%. When MSCs grown and expanded under ambient, or 21%, oxygen, they have different secretory and functional patterns compared to MSCs grown under hypoxic (5%) conditions. Studies also suggest that normoxia-grown MSCs are dysfunctional compared to MSCs grown under chronic hypoxic conditions.

Under normoxic conditions, HIF-1-alpha is rapidly degraded through the ubiquitin system, leading to very low intracellular HIF-1-alpha levels. In the presence of hypoxia, the degradative pathway is shut off. HIF-1-alpha levels increase, forming a heterodimer with HIF-1-beta; this molecule, along with p300, leads to the transcription of multiple genes, including VEGF, FGF, and their receptors, as well as receptors expressed on the cell surface involved in homing, migration, and engraftment—such as CXCR4, CXCR7, and CX3CR1. These receptors home to injured tissues because such tissues express distress signals, i.e., the ligands of these chemokine receptors, including SDF-1 and fractalkine. When stem cells detect tissue distress, they home to that tissue, migrate into it, and engraft. This homing and engraftment mechanism is critical to the healing process. Thus, the hypoxia-induced increase in chemokine receptors should enhance homing and engraftment.

Such data suggest that growing MSCs under chronic hypoxic conditions might result in an MSC with more robust healing properties. In vitro these cells express higher levels of many molecules associated with healing, and the cells are more effective in migrating toward various growth factors and cytokines.

That such in vitro data relate to in vivo activities is suggested by a paper (Tang et al., Circ Res 2009) examining the effects of culturing cardiac-derived progenitor cells for six hours under normoxic versus hypoxic conditions in a mouse model of AMI. The investigators found that cells exposed to hypoxia home and engraft into the ischemic myocardium to a much greater extent than cells exposed to normoxia. Importantly, mice treated with cells exposed to hypoxia had significantly greater recovery of cardiac function than the hearts of mice treated with cells exposed to normoxia.

The Importance of Intravenous Administration of Stem Cells

In studies designed to test the efficacy of cardiosphere-derived stem cells (CDC), such cells were injected into the coronary arteries of pigs four weeks post-AMI (Johnston et al., Circulation, 2009). When 25 and 50 million cells were injected, serum levels of troponin I, a marker of myocardial damage, increased. It was also found that CDCs are fairly large, with an average diameter of over 20 µ—considerably greater than the average diameter of a capillary (10 µ). These findings suggest cell therapy has the potential for coronary plugging, which can lead to further myocardial injury. Wilenski et al., (Circ 2006) showed an image of such a plugged coronary in a pig with AMI undergoing intracoronary injection of MSCs. These investigators warned of the potential for such adverse effects of intracoronary stem cell injection.

In this regard, the lungs are typically considered a barrier to effective delivery of IV injected cells to the heart, where they could exert therapeutic effects. An alternative perspective is that the lungs supply a natural size-filter for IV-administered cells, preventing larger cells from arriving to the heart and plugging coronary arterioles or capillaries. The mean diameter of human MSCs grown under hypoxic conditions appears to be only 11 µ, a finding suggesting they might more effectively bypass the lung barrier when injected IV, and, at the same time, not pose the same risk for intracoronary plug formation as a larger cell would.

If multiple administrations of stem cells are necessary, IV administration provides a convenient, safe and relatively inexpensive strategy. This contrasts with the injection of cells via intracoronary or direct intramyocardial routes, which requires cardiac catheterization and its attendant risks, costs, and inconvenience.


Several new strategies could make adult stem cell therapeutics effective for cardiovascular disease indications. One is the use of allogeneic stem cells derived from young, healthy donors. These are more robust than cells derived from older patients with multiple risk factors. Another is the use of MSCs rather than other cell types, as MSCs are less immunogenic than other cell types. A third is the administration of stem cells grown under hypoxic conditions; this appears to enhance homing, and preliminary evidence suggests they are smaller and, therefore, less likely to cause arteriolar and capillary embolic obstruction. Finally, IV administration of stem cells offers the enormous advantage of allowing multiple dosing conveniently and safely. We are optimistic about these altered stem cell strategies, but we need to await the results of well-designed clinical trials.

Stephen E. Epstein, Ph.D. ([email protected]), is director, translational and vascular biology research at MedStar Heart Institute and clinical professor of medicine at Georgetown University. He also chairs the scientific advisory board for CardioCell, a Stemedica Cell Technologies subsidiary.

Previous articleSmell and Eye Tests May Permit Early Detection of Alzheimer’s
Next articleGSK Picks Up CodeEvolver from Codexis