Review Article

Stem Cell Smart Technology, where are we now and how far we have to go?

Vascular 0(0) 1–13 ! The Author(s) 2017 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1708538117727429 journals.sagepub.com/home/vas

Sherif Sultan1,2, Edel P Kavanagh1,2, Robert Michalus2 and Niamh Hynes1,2

Abstract Approximately eight million people in the United States have peripheral arterial disease, which increases exponentially with age. There have been a plethora of available treatments including surgery, angioplasty, atherectomy, laser technology, and cell-based therapies. Cell-based therapies were developed in the hope of translating laboratory-based technology into clinical successes. However, clinical results have been disappointing. Infusion or injection for stem cell therapy is still considered experimental and investigational, and major questions on safety and durability have arisen. In no option patients, how can they be treated safely and successfully? In this article, we review contemporary practice for cell therapy, its pitfalls and breakthroughs, and look at the future ahead. We introduce a novel smart system for minimally invasive delivery of cell therapies, which exemplifies the next generation of endovascular solutions to stem cell technology and promises safety, efficacy, and reliability. Keywords Peripheral arterial disease, occlusive, smart stem cell delivery system, catheter

Introduction Lower limb peripheral arterial disease (PAD) is a chronic condition that affects an estimated 5% of the population aged 40 years or older.1 Critical limb ischaemia (CLI) represents the most severe stage of PAD and manifests as rest pain, ulceration, or gangrene. CLI is associated with a poor prognosis, with a five-year mortality of 50% and a quality of life comparable to patients with advanced cancer.2,3 An estimated 10% of patients with PAD, older than 50 years, develop CLI within five years.1 Recent reports estimated that the annual incidence of CLI in Europe and the United States ranges between 500 and 1000 new cases per million persons.2 The clinical manifestations and pathological course of PAD are enigmatic and not correlated with the macrovascular disease structure. Although reduced Ankle Brachial Index (ABI) defines the presence of PAD, more than half of patients with a confirmed, abnormally low ABI report no symptoms of PAD.4 The two major, but quite distinct, symptomatic clinical manifestations of PAD, intermittent claudication (IC) and CLI, have very different clinical outcomes.

IC versus CLI Patients with CLI have a six-month risk of major amputation of up to 40% and an annual mortality

of 20%.2,5 By contrast, patients with IC have an annual risk of amputation or death of up to 2%.2,5 Patients with IC progress to CLI at a very low rate (1–2% per limb per year), and a large proportion of patients with CLI do not report antecedent symptoms of IC.1 Furthermore, the degree of the reduction in the ABI does not predict the clinical manifestation of IC versus CLI, or the clinical severity within either group. The fact that more than 90% of those with PAD never progress to CLI makes it apparent that there may be an innately protective mechanism in the majority of those with IC, which defends them from the risk of limb loss. Equally there is something particularly indolent about the minority with CLI that makes them susceptible to limb loss and a priority therapeutic target. Several approaches to management of CLI are available, including risk modification techniques, exercise, 1 Department of Vascular and Endovascular Surgery, Western Vascular Institute, University Hospital Galway, National University of Ireland Galway, Galway, Ireland 2 Department of Vascular Surgery and Endovascular Surgery, Galway Clinic, Doughiska, Royal College of Surgeons in Ireland Affiliated Hospitals, Galway, Ireland

Corresponding author: Sherif Sultan, Clinical Vascular Lead Soalta Hospital Group, National University of Ireland, Galway, Ireland. Email: [email protected]

Vascular 0(0)

2 and pain and ulcer management. However, revascularisation interventions continue to be the cornerstone approach to the treatment of CLI, with variable results on life expectancy, limb salvage, and wound management.3 It is estimated that depending on the stage at which the patient presents and the experience and competency of the treating physician, 50–90% of patients with CLI will undergo some type of revascularisation procedure.1 The advancement of medical therapies and the effect of high intervention rate limit the understanding of the natural history of CLI. Also, many patients might present in advanced stages, and by then, those patients are not fit for surgical intervention. The Society of Vascular Surgery conducted a systematic review and meta-analysis of the natural outcome of those with CLI and found that despite a modest improvement in medical therapies since 1997, one in five patients still die or lose their leg within a year, and wounds and ulcers are more likely to worsen.6 Although the exact reason for the modest improvement is unclear, it is likely related both to improved clinical care of the co-morbidities associated or existing with CLI. This translates to indirect improvements in survival from co-morbidities rather than direct improvement in survival from CLI, and so justifies an attempt at revascularisation whenever possible. The TransAtlantic Inter-Society Consensus for the Management of Peripheral Arterial Disease recommends revascularisation as the optimal treatment for patients with CLI,1 but the type of revascularisation for patients with CLI, whether surgical or endovascular, remains a subject of debate. The evidence comparing the effect of these two main interventions on mortality, morbidity, and limb function remain inconclusive. A recent systematic review and meta-analysis concluded that until more robust data become available, the choice of revascularisation strategy depends on the surgeon’s expertise as well as on the patients’ values and preferences, expected perioperative risk, and anticipated long-term survival.7 Given the broad heterogeneity of this clinical syndrome, future studies should appropriately stratify patients to improve the fidelity of outcomes reporting and the conduct of comparative effectiveness studies in this field. In any case, revascularisation, regardless of the method, does not guarantee limb salvage and even with revascularisation, the odds ratio of limb loss per patient is 1.2 (95% CI 0.6–2.41) within one year. Therefore, a definitive cure for CLI remains an unmet clinical need.

Therapeutic angiogenesis Unlike endovascular therapies, the objective of which is to reopen blocked arteries, cell therapies aim to

promote the growth and development of new blood vessels through the processes of neovascularisation. Neovascularisation involves a coordinated combination of angiogenesis, arteriogenesis, and vasculogenesis. These important events, which occur to some extent physiologically, are influenced by genetic factors but are also the goals of pharmacological and genetic manipulation in therapeutic angiogenesis.

Angiogenesis Angiogenesis involves the sprouting of new capillaries from existing vessels.8 The capillaries are 8–12 mm diameter vessels and lack a developed tunica media. In PAD, at least in the initial stages, the leg muscle is the ischaemic tissue bed from which hypoxia drives angiogenesis. Vascular endothelial growth factor (VEGF)-mediated process of angiogenesis initially results in a large amount of endothelial sprouting that, over time, is followed by ‘pruning’ of the majority of the new capillaries and enlargement of some of the remaining vascular structures. Fibroblast growth factors (FGF) 1 and 2 can cause angiogenesis and, like VEGF, these cytokines work on capillary endothelium via autocrine or paracrine mechanisms.9,10 The ‘new’ endothelial cells created by the process of angiogenesis arise from the proliferation of resident endothelial cells, the recruitment of circulating progenitor cells, or by the conversion of other resident cells into endothelial cells.9,10 The role of skeletal muscle, and its capillary network, in the pathophysiology of CLI, cannot be overestimated. Capillary density in human leg skeletal muscle varies only slightly from person to person and is largely stable over time. However, it can increase with activities such as structured exercise, or decrease via a process called vascular rarefaction during periods of marked reductions in muscle use, or in the setting of conditions such as hypertension. In PAD blood supply to the skeletal muscle in the calf is compromised, and the resulting ischaemia-induced angiogenesis can have important effects on a patient with PAD. First, and most simply, ischaemia-induced angiogenesis in the leg muscle distal to the arterial blockage can reduce resistance and allow some additional perfusion through existing collateral vessels.5 The effects of PAD on capillary density in leg muscle, however, are more complicated because, if ischaemia-induced angiogenesis were the only process to consider, one would expect that capillary density would be higher in the leg muscle of patients with PAD than in age-matched control individuals. By contrast, capillary density is, in fact, lower in patients with IC than in age- and sex-matched control individuals.11,12 Therefore, processes such as

Sultan et al. ongoing cell death could be opposing angiogenesis and resulting in lower capillary density in the leg muscle.13 In patients with IC, the lower the capillary density in the most symptomatic limb, the worse the patient’s functional performance. After supervised exercise training, angiogenesis occurs in the calf muscle before changes in functional performance and, in turn, changes in functional performance occur without changes in leg blood flow or ABI, indicating a central role for the skeletal muscle microvasculature.14 In patients with CLI, bypassing arterial occlusions into downstream arterial vessels of 1–4 mm diameter is often ineffective, and one possible explanation could be defects in the vasculature distal to the graft.15 Together, these factors indicate that modification and manipulation of the capillary bed in ischaemic skeletal muscle microvasculature far downstream of an arterial occlusion could have the potential to enhance the clinical course of PAD.

Arteriogenesis Arteriogenesis involves the formation of vessels, which are 20–100 mm in diameter and have a fully developed tunica media, which allows vasodilation and regulation of blood flow. Arteriogenesis is likely to involve the utilisation and subsequent maturation of pre-existing collateral vessels, as well as the de novo formation of mature vessels.16 Unlike angiogenesis, ischaemia is not a prerequisite for arteriogenesis.17 The primary arteriogenic stimuli include shear stress and inflammation, which results in an invasion of white blood cells. Monocytes and macrophages surround collateral vessels using cytokine-specific and cell-specific mediators,17,18 and produce growth factors, such as FGFs, and tumour necrosis factor. This process leads to downstream signalling through syndecans with subsequent vascular remodeling.18 Angiogenesis and arteriogenesis can occur independently of each other, but they can also occur together. Angiogenesis is likely to be the initial event, and the synergy between the processes is superior to either alone.

Vasculogenesis Vasculogenesis defines the in situ formation of new blood vessels by aggregation of primitive endothelial precursor cells. Vasculogenesis was once thought to be limited to embryonic development. However, in the late 1990s, Asahara et al.10 showed that new blood vessels that form following ischaemia contain endothelial progenitor cells,9 which can be involved in both angiogenesis and arteriogenesis. In ischaemic tissue, endothelial progenitor cells are capable of differentiating into endothelial cells and have the ability

3 to promote neovascularisation. Progenitor cells can be circulating bone marrow-derived cells that incorporate into blood vessels, or cells from other sources that hone to developing vessels where paracrine mechanisms promote vessel growth.19,20 Stromal cell-derived factor-1 is a cytokine involved in vasculogenesis and vascular remodelling, but many of the other elements in this process remain poorly understood.

Clinical applications for neovascularisation Recognition of some of the deficient mechanisms in neovascularisation experienced by those with CLI has presented researchers with the opportunity for therapeutic targets. Proof of the therapeutic angiogenesis concept was demonstrated in animal models. These resulted in clinical trials of pro-angiogenic growth factors, delivered either as proteins or as genes that encode them,21 and/or cell therapy, e.g. endothelial progenitor cells or other angiogenic stem cells,22 to stimulate and promote angiogenesis in ischaemic tissues. The clinical trials have confirmed feasibility and short-term efficacy of intramuscular injection of these biologic therapies, and safety has been demonstrated by a lack of ‘offtarget’ angiogenesis, no growth of occult tumours, or no progression of diabetic retinopathy. Examples of preliminary clinical successes include intramuscular injection of hepatocyte growth factor plasmids, which improved blood perfusion in CLI patients. The enhanced perfusion was demonstrated by ABI increase from 0.46 to 0.59 and a greater than 25% reduction in ischaemic ulcer area.23 Clinical improvements in CLI patient symptoms were also reported upon intramuscular injection of both autologous bone marrow (BM) mononuclear cells and VEGF plasmids. This intervention led to improved perfusion (ABI increased from 0.26 to 0.49) and reduction in rest pain.24 Despite promising initial results, some recent phase II and phase III clinical trials of angiogenic gene therapy did not generate consistently expected benefits.25 The strategy of direct injection of naked plasmids carrying angiogenic genes was found to be ineffective due to low cellular transfection efficiency. Using viral vectors to deliver the genes can overcome the low transfection efficiency. However, the viral vectors raise safety concerns due to random insertion into the host cell genome, and the risk of integration of genes that lead to a persistent expression of angiogenic factors with consequent risk of pathological angiogenesis or tumorigenesis.25 Recombinant proteins are less likely to cause such long-term safety issues. However, growth factors have short circulation half-lives, requiring multiple injections to achieve sufficient and

4 sustained growth factor levels at the ischaemic site. Multiple doses of angiogenic growth factors may cause adverse effects such as hypotension,26 vascular leakages,27 and tissue oedema.28 Current cell therapy methods are also facing some potential challenges such as low cell retention, low viability post-transplantation, and limited integration into host tissue.29 Trials of therapeutic angiogenesis in humans with CLI have almost consistently involved stand-alone strategies in patients who have no option, or very limited options, for revascularisation. Although angiogenesis is the growth of blood vessels, to meet the goals of improving clinical outcomes for patients with CLI, a long bridge of vessels extending from the thigh to calf, or lower, would be needed to increase blood flow to the distal leg. This is not realistic, especially in the time frame required for limb salvage. Therapeutic genes, such as VEGF, which have been disappointing in the treatment of PAD, have been successful and even transformational when VEGF served as the target for inhibition in the treatment of cancer and macular degeneration.30,31 The simplest reason for this difference is that data from human studies of therapeutic angiogenesis have failed to compellingly demonstrate evidence of successful gene transfer following intramuscular delivery, even when changes in systemic measures of the transgene or its product should have been possible.32,33 Gene expression following gene transfer with an intramuscular delivery of therapeutic genes in humans is often localised to the injection site and does not extend to large areas of ischaemic muscle in the thigh and calf. Also, investigators in most trials make ‘blind’ injections without ultrasound guidance, and this often results in the agent being delivered to fascia or subcutaneous tissue rather than to skeletal muscle.34 Even with the correct agent and successful vector delivery, optimistic estimates of transfection efficiency in humans are only in the single-digit per cent of cells for either plasmid- or adenovirus-mediated gene transfer.35 Therefore, with the plasmid or adenoviral vectors that were used, the expected magnitude and duration of gene expression are very limited and subtherapeutic. To further complicate matters, the optimal choice of muscle to inject and the optimal injection site within that muscle are still unknown.

Angiogenesis in ischaemic muscle Skeletal muscle is the major metabolically active tissue in the legs by its mass, and plausibly provides paracrine or autocrine signals to influence the response to ischaemia. In humans, the pattern of change in gene expression in ischaemic skeletal muscle depends on whether or not the patient has CLI or IC.36 Patients with PAD

Vascular 0(0) have an angiogenic response to repetitive or constant ischaemic stimuli in the distal calf muscle. However, in individuals with CLI, angiogenesis does not lead to the remodelling of the larger vessels that is needed to promote tissue perfusion. The reasons for this defect are still largely unknown. In chronic or repetitive ischaemia, myocytes in the skeletal muscle like cardiomyocytes have changes in the expression of genes protective against critical processes such as apoptosis and the effects of reactive oxygen species.37,38 The endothelium in ischaemic muscle might be especially important. In humans, capillary density in the most symptomatic areas of calf muscle in patients with PAD correlates with functional performance measures.14,39 In addition, changes in calf muscle capillary density occur before changes in functional performance in patients with PAD following exercise training.14,39 Together, these data strongly support the hypothesis that cells in distal ischaemic skeletal muscle have an important role either in a patients’ functional capacity or their response to therapy in PAD. Therefore, factors that can modulate the capillaries in ischaemic muscle in PAD could serve as goals for therapeutic angiogenesis, but perhaps not as stand-alone therapies. Circulating BM-derived progenitor cells are dysfunctional and levels are lower in patients with CLI than in healthy controls because of prolonged proinflammatory stimuli.40 These low rates might explain the absence of treatment effect seen in all clinical trials. However, Mac Gabhann et al. suggested that this disease-mediated cell dysfunction in patients with CLI is reversed in BM-derived mesenchymal stem cells (BMMSCs),41 making them possibly more efficient in neovascularisation therapy for CLI. Also, BMMSCs might provide additional benefit when used in an allogeneic approach, for example off-the-shelf availability and circumvention of BM aspiration procedures in the frail patient with CLI.42 Although a clear difference between trials that used intra-arterial and intramuscular administration was not observed, evidence shows that BM-derived cells mainly act via paracrine pathways. One advantage of intramuscular administration could be the creation of ‘local depots’ of stem cells with increased local paracrine activity and local release of arteriogenic cytokines.10,43 This suggests that muscle leg mass utilisation has to be explored in future clinical trials. A meta-analysis of 10 placebo controlled trials of BM-derived cell therapy in 499 patients with CLI showed no advantage of cell therapy on the primary outcome measures of amputation, survival, and amputation-free survival (AFS).44 This meta-analysis underlines the need for future well-designed doubleblinded and placebo-controlled randomised trials that

Sultan et al. are adequately powered, to investigate specific BMderived cell therapeutic strategies. From the available evidence, it is believed that future cell therapy in CLI should focus on specific cellular therapies. It has recently been seen that the neovascularisation capacity of mesenchymal stem cells is not compromised in patients with CLI. This capability suggests that autologous mesenchymal stem cells may be suitable for cellular therapy in patients with CLI.8 When combining the results of 12 RCTs in which >500 CLI patient were treated, a significant benefit was demonstrated (P

Stem Cell Smart Technology, where are we now and how far we have to go?

Approximately eight million people in the United States have peripheral arterial disease, which increases exponentially with age. There have been a pl...
427KB Sizes 2 Downloads 9 Views