Can Angiogenic Doping Boost Athletic Performance?

Can angiogenic doping increase athletic performance? While delivering VEGF-A results in an angiogenic response, it has been shown not to increase blood supply in healthy tissues that already have an adequate blood supply. VEGF-A gene therapy effects are more pronounced in tissues with impaired blood flow, such as ischemic tissues [200, 216]. Studies have shown that overexpressing or overdelivering VEGF-A leads to the formation of a chaotic, leaky, and irregular blood vessel network that is incapable of upgrading the vascular system [153]. It is not entirely clear how a sophisticated vascular patterning is achieved, which optimizes blood oxygen uptake and delivery, but growth factor gradient formation, exploiting the differential affinities of the multiple VEGF-A isoforms, is thought to play a central role [318, 319] (see Fig. 6).

The treatment of ischemic diseases with VEGF-A has shown that many hurdles have to be overcome to achieve clinically relevant results [200]. However, in highly ischemic tissues of coronary artery disease patients, impaired blood flow is better than none. It has also been speculated that VEGF-A alone may not be sufficient to improve blood flow, and that a combination of angiogenic growth factors in appropriate ratios might be necessary [344]. For example, combining VEGF-A with angiopoietin-1 could reduce the leakiness of newly formed vessels, leading to more physiological outcomes [344, 345].

Systemic Versus Local Delivery and the Importance of Gradients

The current view is that effective direct VEGF application requires complex and tightly controlled delivery methods, making it largely impractical for doping attempts currently. Systemic administration of larger amounts is considered hazardous due to the vasodilatory effect of VEGF-A observed in early clinical trials and animal studies [215, 346]. Systemic administration of VEGF-A might not have any positive impact on angiogenesis. In one study, intravenous delivery of a VEGF-A form engineered for prolonged half-life actually resulted in a reduction in capillary density in the kidney target organ [347].

Hence, as a protein, VEGF-A can presently only be applied locally. However, the distribution of locally applied VEGF-A does not replicate the endogenous growth factor gradients necessary to organize angiogenic sprouts and ultimately result in a hierarchical, functional network. In the absence of these instructive gradients, VEGF-A application typically results in nondirectional growth, which does not necessarily improve oxygen and nutrient supply (see Fig. 4B). For example, topically applied VEGF-A in the chorioallantoic membrane assay results in extensive angiogenesis, but not in a hierarchical, streamlined network [24, 25]. The early clinical trials to increase vascularization of ischemic hearts using VEGF-A had no clinically meaningful effects, at least partly for the same reasons, despite measurable angiogenesis [200].

Angiogenic Doping—without Genes and without Increasing VEGF Levels

The goal of angiogenic doping is to increase vascular density slightly and uniformly across large areas of the body, particularly in skeletal muscle. This goal has been achieved, but with lymphatic vessels rather than blood vessels. Kataru et al. increased lymphatic vascular density in mice throughout the body without altering VEGF-C levels [348]. This increase was achieved by conditionally inactivating a gene that negatively regulates VEGF-C-mediated intracellular signaling in lymphatic endothelial cells. The same paradigm should work equally well in blood vascular endothelial cells with VEGF-A signaling. Kataru et al. used genetic engineering to implement ‘molecular nudging’ of the PTEN gene, but pharmacological inhibitors of PTEN or other phosphatases could enable targeting of the intracellular signaling of VEGFR-2 [349]. Nobody has been searching intensively for suitable compounds, as the primary goal of the pharmaceutical industry was the opposite: to find inhibitors of intracellular signaling downstream of the VEGF receptors, such as receptor tyrosine kinase (RTK) inhibitors [350]. Protein tyrosine phosphatase (PTP) inhibitors, on the other hand, target the off switches for RTK signaling. VE-PTP and PTP1b are examples of endothelial cell-specific PTPs [351, 352]. The VE-PTP inhibitor razuprotafib has been tested in clinical trials, but nothing is known about its effects on sports performance [353, 354].

Small-molecule drugs such as PTP and PHD inhibitors could also be targeted to the vasculature. Targeting pericytes or vascular smooth muscle cells, which are natural sources of VEGF-A, could be done with an antibody conjugate directed against platelet-derived growth factor receptor-β or other mural or endothelial cell surface markers. Targeting would allow for a sufficient local concentration despite almost undetectable systemic levels.

Current Technical Barriers and Suspected Current Adoption

Based on the technical limitations of delivery described in Sects. 4.1 through 4.5, it appears unlikely that currently available pro-angiogenic gene therapies, CRISPR or mRNA-based drugs, will be misused for doping purposes in the near future. However, the same cannot be said for erythropoietin delivery, as microdosing of mRNA drugs appears fully feasible with currently available agents.

The leading VEGF candidate for future misuse would likely be VEGF-D, not only because of its superior biological properties, but also because many fewer methods exist for its detection, as it is the most recently discovered and arguably least researched VEGF. Very different from VEGFs are hypoxia mimetics and phosphatase inhibitors, which are currently available and easy to administer. Microdosing of hypoxia mimetics, including CO, could produce a small but meaningful effect on erythropoiesis, which might go undetected due to the concurrent effect on angiogenesis and blood volume expansion. From a detection point of view, CO is very similar to high-altitude training and indistinguishable from environmental exposure [355]. Similarly, systemically microdosed phosphatase inhibitors could slightly increase the response to normal levels of VEGF-A. Moreover, the development of small-molecule analogs is well understood [356], and current detection efforts do not even cover all known compounds.

While all such pharmacological interventions carry significant risks, moderate high-altitude hypoxia, in contrast, appears to be cardioprotective in both animal models and humans [357, 358]. However, the mechanisms underlying protection remain unclear because the multitude of concurrent physiological changes associated with high-altitude exposure creates uncertainty about which changes are causative [359]. Additionally, similar benefits may be achieved through different adaptive strategies, as genetic variation shows [360]. In the absence of high-altitude hypoxia, aerobic exercise is the most effective way to achieve a similar effect [361]. Table 2 summarizes the existing and potential methods to stimulate angiogenesis for doping purposes, along with their detectability and practical considerations.

Agent | Example(s) | How to detect it | (Likelihood of) current use | Remarks: ease of synthesis, acquisition, and application
(Simulated or real) altitude living/training | Hypoxic tents, hypobaric chambers | No need for detection, as it is not banned | Common | Easy
Iron competitors and chelators | Carbon monoxide, CORMs, deferoxamine, Co2+ | Spectrophotometry based on the different absorption maxima of carboxyhemoglobin vs other hemoglobin forms; smokers always test positiveThe iron chelator deferoxamine can be detected by mass spectrometry or HPLC, but has a very short half-life [253] | High | Readily available, correct dosing is challenging for carbon monoxide
PHD/HIF inhibitors | Roxadustat (FDA-approved), IOX5 (ongoing clinical studies) | Different types of mass spectrometry | Medium (new compounds are being developed) | Readily available, oral administration
PTP inhibitors | Razuprotafib (AKB-9778) | Different types of mass spectrometry | Medium | Compounds are commercially available. Failed in clinical trials to treat, e.g., diabetic nephropathy and proliferative diabetic retinopathy, but might still work for doping
Targeted small-molecule drugs | Antibody conjugates of HIF stabilizers and hypoxia mimetics | Difficult to detect due to small amounts and targeting | Medium–low | Requires custom synthesis and injection
Gene therapy | AdVEGF-D, AdHIF-1α | PCR (within a few weeks of use) | Low | Requires expensive and specialized equipment for production and administration
Growth factor proteins | VEGF-A, VEGF-D | Western blot (within a few weeks of use) | Low | Currently, it remains challenging to deliver the agent in a way that benefits athletic performance
mRNA | VEGF-A, VEGF-D, HIF-1α | RT-PCR (within a few weeks of use) | Low | Currently, it is technically challenging to deliver the agent in a way that benefits athletic performance
Gene editing, e.g., CRISPR-based drugs | SNP-editing of EPO or VEGFA promoters to slightly increase the endogenous hormone levels | Depending on the delivery system, (RT-) PCR could be used within a few weeks of application. Mosaicism resulting from incomplete editing is likely to remain undetectable for years to come | Low, but rapidly increasing | Delivery of the CRISPR drugs is still a high technical hurdle, but it might be easier to achieve for doping than for therapy. Target cells can be muscle cells, muscle-derived stem cells, or satellite cells
miRNA | miR-126, miR-210 | Dependent on the delivery system. Targeted delivery (e.g., aptamers or customized extracellular vesicles) can be challenging to detect | Very low | Requires expensive and specialized equipment for production and administration
Slow-release biomaterials | Hyaluronic acid hydrogels, poly- (lactic-co-glycolic acid), various nanoparticles | Slow-release materials are relatively easy to detect because of their long biological half-lives | Very low | It can be used in combination with any of the above methods