Gene Therapy for The Prevention of Vein Graft Failure: Are We Getting Close?

Michael S. Conte MD

Vein bypass grafting has proven to be a versatile, effective and durable treatment for coronary and peripheral atherosclerosis. Nonetheless, vein graft failure is a common clinical occurrence that results in significant morbidity and mortality. Some 40% of infrainguinal vein grafts will either occlude or require revision within 5 years, and re-do procedures are marked by increased technical complexity and inferior long term results.

 Vein grafts fail most commonly due to the development of intimal hyperplasia (IH), a flow restricting lesion that may occur diffusely throughout the graft or, more commonly, at focal sites near anastomoses or within the body of the graft. This process, which is responsible for the majority of mid-late term (3-24 month) failures, is poorly understood and its prevention has been a subject of intense investigation. While our knowledge of the pathophysiology of vein graft failure remains incomplete, there are several potentially relevant pathways that may provide therapeutic targets in the healing graft including: thrombosis, inflammation, ischemia/reperfusion, proliferation, apoptosis, matrix metabolism, and lipid accumulation.

 To date, outside of the modest benefits associated with antiplatelet therapy,  pharmacologic approaches to prolong vein graft patency have not borne fruit. Genetic approaches to modulate graft healing are actively being studied both experimentally[i] and clinically, and hold significant promise for the future (3).Vein bypass grafts are uniquely amenable to genetic manipulation because the delivery of genetic material can be achieved under carefully controlled, ex-vivo conditions which favor both safety and efficiency. Genetic interventions performed at the time of graft implantation offer the potential to alter the long term healing response that ensues.

Genetic interventions may take several forms. In its purest sense, the commonly used term ‘gene therapy’ implies the delivery of a fully active transcriptional unit (gene) to cells of the body, to replace or augment the expression of its protein product in either a local or systemic fashion. Genes are large DNA molecules that are inefficiently taken up by cells and therefore a delivery system (‘vector’) is required to achieve meaningful levels of expression in a target tissue. Achieving gene transfer to a majority of cells within a vein graft is a significant hurdle, particularly within the temporal constraints of an intraoperative strategy. The ‘ideal vector’ for this purpose has yet to be defined but would have the following desirable attributes: 

·         Safe- eliciting minimal or no host inflammatory response or toxicity

·         Easily produced, handled, and stored in clinically relevant concentrations

·         Capable of rapid and efficient gene delivery to non-dividing cells

·         Confer stable or regulatable levels of gene expression

·         Flexible to accommodate genes of all sizes 

For cardiovascular applications, most vector development efforts have focused on the use of re-engineered viruses, most notably adenovirus and adeno-associated virus (4). Non-viral systems are also being investigated, but significant limitations exist in their current state. Though considerable progress has been made in recent years, and clinically usable systems are already available, further development of vector technology will be required to achieve a method for intraoperative gene transfer to vein grafts that fulfills these criteria.

 While the technical obstacle of a usable vector system may be surmountable in the near future, an equally vexing question remains that of which gene or genes might best promote long term vein graft function. Numerous potential candidate genes have been identified in animal studies targeting the various processes listed above. Genes encoding proteins that result in cell cycle inhibition, diminishing the cell proliferation that characterizes IH, have received the greatest attention.

 A further attractive property for a candidate gene, given the delivery limitations noted above, would be the ability to affect surrounding cells. In this context, genes encoding secreted proteins or enzymes generating diffusible products may produce a significant local effect even if only a minority of the cells in a given region of the graft are transduced. The most notable example would be the nitric oxide synthase (NOS) isoforms, which are leading candidates for cardiovascular gene therapy. Indeed, NOS gene therapy has reached the stage of clinical testing in coronary angioplasty and we will likely see other peripheral applications in the near future. Further preclinical investigations in animal models of vein grafting will be important to elucidate the most relevant genes for subsequent clinical trials.

Another approach to genetic manipulation, which does not require the efficient transfer of large, intact genes, is gene inhibition. A specific gene, or an entire cellular program (e.g. cell cycle), may be inhibited using small nucleic acid molecules- oligonucleotides (ODN)- that may function to either block transcription (‘antisense’) or block the activity of critical proteins (known as transcription factors) that control gene expression. The latter strategy involves the design of small, double-stranded DNA molecules that serve as a ‘decoy’ for the transcription factor, preventing it from interacting with its normal sequence target in the cell’s chromosomal DNA.

  A major attraction of these gene inhibition strategies is that these small oligonucleotides may be delivered far more easily to cells and tissues with high efficiency, and do not require specific viral or non-viral vectors. The use of non-distending pressure (300 mm Hg) has been shown to result in >80% uptake of oligonucleotide by cells within the saphenous vein wall within ten minutes of exposure.

 A ‘decoy’ approach targeting cell proliferation was recently developed and tested in a rabbit model of vein bypass grafting (5). A double stranded oligodeoxynucleotide (14 base-pairs) was designed to incorporate the binding site for the transcription factor E2F, which controls the expression of multiple genes that are responsible for progression of the cell cycle in proliferating cells. Vein grafts treated with the E2F decoy in solution at the time of implantation demonstrated marked inhibition in intimal hyperplasia and graft atherosclerosis for up to 6 months in cholesterol fed rabbits. The grafts appeared to have altered remodeling with the development of medial hypertrophy in place of intimal hyperplasia, and also demonstrated improved endothelial function (6).

 Based on this preclinical data, a pilot clinical trial (PREVENT I) was carried out at a single institution in patients undergoing lower extremity bypass with autologous vein (7). A total of 41 patients were prospectively randomized, in double-blind fashion, to placebo (saline, N=16), E2F decoy (N=17), or a scrambled sequence ODN (N=8). Intraoperatively, the veins were harvested and passed to a back table where they were mounted on a cannula and inserted into a device for pressure-mediated transfection with ODN. This device bathes the graft in ODN solution on all sides, allowing for uniform distribution of a pressure of 300 mm Hg without distension of the graft. Transfection occurred at room temperature for 10minutes, after which the graft was passed back to the operative field. Small residual segments of the treated vein were taken back to the laboratory and revealed successful delivery of the ODN to 90% of cells in the wall, with significant inhibition of the target cell cycle genes.

 There were no systemic complications or serious adverse events noted in the ODN treated patients. Postoperative morbidity was similar for all groups. The percentage of ‘high-risk’ grafts was notably high (59%) in this study, and was evenly distributed among treatment groups. A high rate of graft revision was observed in the untreated grafts, accounting for a marked reduction in the overall event rate at 12 months in the E2F decoy- treated patients (29% vs. 69% placebo, p<.05), Attesting to the efficacy of surveillance and revision, assisted primary patency rates were similar at 88% and 81% respectively. Most intriguingly, there was a marked reduction in ‘non-technical’ failures (> 1month) among patients with high-risk grafts receiving E2F decoy treatment (1/7 vs. 6/6 untreated), suggesting that those grafts at highest risk for IH demonstrated the greatest benefit. This study demonstrated the safety and feasibility of intraoperative transfection with the E2F decoy ODN, and suggested the possibility of biological efficacy. Recently, a corporate-sponsored phase II trial of E2F decoy treatment of coronary vein grafts has been completed in Germany with 200 patients enrolled and randomized. The final results, including repeat angiography as an endpoint, have not yet been released.

 Based on these results, a large scale randomized, controlled trial of E2F decoy ODN treatment for the prevention of lower extremity vein graft failure has been initiated by the corporate sponsor, Corgentech (Palo Alto, CA). This trial is designed as a Phase III, multi-center, randomized, placebo-controlled study in patients undergoing infrainguinal vein grafting for critical limb ischemia. Patients will be randomized to intraoperative treatment of the vein graft with the E2F decoy versus control (saline), delivered by the non-distending pressure method. The primary endpoint will be a composite one-year incidence of reintervention for graft failure (stenosis or occlusion) or limb loss. Approximately 1400 patients will be randomized at some 40-60 centers across North America. This trial is likely to be a model for future studies of genetic interventions in vein grafting, and represents an exciting opportunity to address a critical problem with a targeted molecular therapy performed by surgeons in the operating room. These recent developments herald the beginning of an era of genomics-based interventions in vascular disease, and many of the earliest applications are likely to involve surgeons and interventionalists.