In many ways, erythrocytes would make an ideal agent for drug delivery. They are the most abundant cellular constituent in blood, and on average travel hundreds of kilometres in the vasculature during their lifespan (100–120 days). If they could be converted into carriers, they could, in theory, prolong the circulation and improve the bioavailability of drugs that act in the bloodstream. Loading therapeutic agents into erythrocytes is of limited value, as many drugs diffuse poorly across the erythrocyte membrane. However, coupling agents to the erythrocyte surface could overcome this problem. Now, Muzykantov and colleagues provide the first proof of principle for this drug delivery strategy in Nature Biotechnology by showing that this can improve the therapeutic profile of tissue plasminogen activator (tPA) in prophylaxis of vascular thrombosis (a blood clot that develops in an artery and remains in place).

Existing treatments that lyse clots (fibrinolysis), such as tPA, are good for treating acute post-thrombotic ischaemic events, such as myocardial infarction, but are unsuitable for prophylaxis because they have a short lifespan, dissolve both pre-existing and nascent clots (ideally, nascent clots should be selectively targeted), and leak out of the vasculature and cause side effects in other tissues. But biotinylating tPA and erythrocytes, and crosslinking them using streptavidin, produces a stable complex that is a potent thrombo-prophylactic agent in animal models of venous and arterial thrombosis. The erythrocyte–tPA complex extended the duration of tPA's fibrinolytic activity and was ten times more selective in lysing nascent over pre-existing pulmonary emboli (blood clots that break off, travel through the bloodstream and stick in a pulmonary artery) than free tPA, and twenty times more selective in lysing nascent over pre-existing arterial clots. Prophylactic injection of erythrocyte–tPA dissolved occlusive thrombi in the carotid artery soon after their formation, even though they were resistant to fibrinolysis by free tPA injected before or after thrombosis.

The authors say that the strategy works because the coupled tPA is denied access to the relatively impermeable preformed clots and extravascular tissues, thereby reducing the risk of bleeding. The complex circulates for prolonged periods and becomes incorporated inside nascent clots, initiating lysis from within (which they call “the 'Trojan horse' fibrinolysis of nascent clots”). Proving that the method works in other preclinical and clinical models will be important, but a more interesting question will be the wider one of how many other agents can have their therapeutic profiles enhanced by coupling to erythrocytes.