Further evaluation needed: Therapeutic angiogenesis in the treatment of PAD


Lars Norgren talks to Vascular News about therapeutic angiogenesis in the treatment of peripheral arterial disease (PAD). He points out that further evaluation in this area is of great importance and that biomarkers may assist in the understanding of who might respond to the therapy.

The history of therapeutic angiogenesis in PAD is young. As late as 1996, Jeffrey Isner and coworkers were able to show an increased collateralisation by applying plasmid DNA encoding for VEGF1 to a patient with critical limb ischemia (CLI). Iris Baumgartner was the first to present a patient series according to the same principles, the plasmid being injected intramuscularly.2 The first series treated with autologous bone marrow- or peripheral blood harvested mononuclear cells was presented by Tateishi-Yuyama in 2002.3 Besides improved collateralisation, anti-inflammatory effects are crucial and regenerative processes are thought to be of importance. However, gene- and cell therapy studies have run in parallel during a 20 years period without any evident breakthrough. The great majority of trials have included patients with CLI, some few recruited subjects with intermittent claudication.

Gene therapy

Various fibroblast growth factors have been used, specifically non-viral 1FGF. The TALISMAN phase II study4 concluded a reduction of major amputations, while the larger phase III trial, TAMARIS,5 could not reproduce this finding. The Hepatocyte Growth Factor (HGF) has in small trials exerted effects on symptoms as well as on hemodynamic parameters.6,7 The planned larger trial, AGILITY, had to be cancelled due to slow recruitment of suitable patients. A Cochrane analysis8 of relevant gene trials concluded bias and  imprecision and in general inconsistant findings in the studies.

Cell therapy

Numerous studies have been published, the majority utilising autologous bone marrow mononuclear cells (BM-MNC), injected intramuscularly. Some few studies harvest cells from peripheral blood, requiring stimulation by granulocyte colony stimulation factor (G-CSF). Single studies have used intraarterial administration of the cells. A number of studies have found a reduced major amputation rate in the cell treated group compared to controls. Most importantly, however, three metaanalyses have concluded that in placebo controlled trials, a significantly reduced major amputation rate has not been achieved in the treatment group compared to the control group.9–11

Autologous cells should be the reasonable choice to avoid immunological reactions. It has, however, been found that progenitor cells in older individuals are both limited in numbers, less effective and  the rate of endothelial progenitor cells is inversely correlated to the Framingham risk.12 Furthermore, progenitor cells of certain importance (CD34+) are less common in CLI patients.13

Allogeneic cells, umbilical cord derived, bone marrow mesenchymal stromal cells from healthy volunteers as well as placental mesenchymal-like expanded adherent stromal cells (PLX-PAD) have been or are under evaluation, mostly in CLI patients, but PLX-PAD cells have also been used in one intermittent claudication study. Placental cells imply a young source with immune privileged characteristics, higher capacity in releasing proangiogenic proteins and downregulating the inflammatory process. The claudication PLX-PAD trial showed that injection of cells from two different donors at the two administrations was of benefit,14 a fact transferred to the Phase III trial in CLI, PACE.15

Patient inclusion

The majority of trials use the inclusion criterion ”no-option” (for revascularisation), based on a rate of about 40% such cases in the CLI population.  Despite that, recruitment to the studies has been very slow, indicating ”no-option” as rare. An evident reason is the gradual increase of revascularisation, but also trial based exclusion criteria, and the fact that patients in end-stage CLI should not be included as they are less likely to respond to treatment, and the risk of loss to follow up is considerable. For that reason a ”poor option” criterion has been adopted in more recent trials, indicating that although a revascularisation is not excluded, there is an evident risk for technical failure or medical complications. There are also patients with failed revascularisations, who might have more stable symptoms and do not require an immediate amputation, and therefore appropriate for trials.

Primary efficacy endpoint

A considerable interpretational problem exists in the fact that a mix of events are used for the primary efficacy endpoint. In the CLI trials amputation-free survival (AFS), ulcer healing (percentage, complete healing), change of pain, improved hemodynamics, number of collaterals and composite endpoints are used to describe a ”positive” event.  Reversely, doubling of ulcer area, appearence of de novo gangrene together with ”first occurence of treatment failure” are used.

Why no evidence yet?

There are several potential reasons why the therapeutic angiogenesis approach for CLI has not yet achieved evidence for or against its use.  The variation in study design is important. Among seven active CLI studies, registered in ClinicalTrials.gov, 4 are utilizing bone marrow cells, compared to either placebo or autologous peripheral blood, 3 allogeneic product studies are registered, using placebo controls.  Two studies are delivering the cell product intra-arterially, four use an intramuscular delivery while one uses both routes. Whether the delivery of the angiogenic product should take place above or below vascular lesions, occlusions or stenoses, is not clear. If the great variation in this localisation and in the number of injections and injection sessions is of importance is not known.

What is measured? Is the response of clinical value? First, the considerable difference between the Rutherford 4 and 5 categories is important. When only rest pain is the symptom to be quantified, the problem with this measurement has to be considered. The clinical value of pain relief is, on the other hand, important. Lesions can be evaluated, new lesions can be recorded, while measuring the grade of healing gives rise to problems. Regarding the hierarchy of outcomes, keeping the limb should be the primary goal for the patient, and for that reason amputation-free survival seems the most reasonable primary efficacy endpoint. Although complicated, the grade of ulcer healing may be a valid secondary endpoint. Hemodynamic parameters create certain problems as perfusion can not be supposed to increase considerably. Both ankle-brachial index, toe pressure (possibly toe-brachial index) and tcpO2 deserve to be followed, but not used as primary efficacy endpoints. The value of PET-CT and NIRS (near infrared spectroscopy) and other methods has to be evaluated to get an appropriate measure of perfusion.

Most importantly, biomarkers related to CLI, which possibly can be expected to normalise during an effective treatment should have a place in this setting. It has been shown that pro- and anti- inflammatory cytokines (e.g. IL-6/IL-10) and growth factors (e.g. VEGF) may be modified in relation to clinical improvement. Further evaluation is of great importance as biomarkers may assist in the understanding of who might respond to the therapy.

What should be done?

When including patients: look for failed revascularisations and poor option cases !  Perhaps a risky procedure can wait ! If a procedure fails, and an amputation is not immediately required, there is a second possibility to include the patient ! Avoid end stage CLI patients, as the chance of relevant follow up and  success is minimal. When designing a trial: homogeneity is crucial to enable comparison and interpretation. This relates to patient inclusion and to selection of the primary efficacy endpoint. Amputation-free survival seems most relevant. The selection of cell population (or gene product), delivery route, number and sites of injections may depend on the characteristics of the delivered substance, however, intramuscular injections seem mostly relevant. The use of young (allogeneic) cells seems of great interest.

An obvious question arises, as there seems to be few patients suitable for studies, is the therapy actually required? Most likely it is, as facilities for complicated revascularisations are not accessible everywhere, resources do not exist in low-and middle income countries, and both ”no option”, ”poor option” and failed revascularisations are realities. Focus should be on CLI, the need for angiogenic treatment of intermittent claudication seems more doubtful. Further studies, evaluating this kind of therapy adjuvant to revascularisation are also awaited.


  1. Isner JM, Pieczek A, Schainfeld R et al. Lancet 1996; 348:370
  2. Baumgartner I, Pieczek A, Manor O et al. Circulation 1998; 97:1114
  3. Tateishi-Yuyama E, Matsubara H, Murohara T et al. Lancet 2002; 360:427
  4. Nikol S, Baumgartner I, Van Belle E et al. Mol Ther 2008; 16:972
  5. Belch J, Hiatt WR, Baumgartner I et al. Lancet 2011; 377:1929
  6. Powell RJ, Simons M, Mendelsohn FO et al. Circulation 2008; 118:58
  7. Powell RJ, Goodney P, Mendelsohn FO et al. J Vasc Surg 2010; 52:1525
  8. Forster R, Lieuw A, Bhattacharya V et al. Cochrane Database Syst Rev 2018; 10:CD012058
  9. Teraa M, Sprengers RW, van der Graaf Y et al. Ann Surg 2013; 258:922
  10. Peeters Weem SM, Teraa M, de Borst GJ et al. Eur J Vasc Endovas Surg 2015; 50:775
  11. Pan T, Wei Z, Fang Y et al. Vasc Med 2018; 23:219
  12. Hill JM, Zalos G, Halcox JP et al. N Engl J Med 2003; 348:593
  13. Teraa M, Sprengers RW, Wetserweel PE et al. PLoS One 2013; 8: e55592
  14. Abstract AHA 2018
  15. Norgren L, Weiss N, Nikol S et al. Eur J Vasc Endovasc Surg 2019; 57:538

Lars Norgren is Emeritus Professor of Surgery at Örebro University in Sweden.


Please enter your comment!
Please enter your name here