Lab-grown blood: how close is it?

What lab-grown blood is

Red blood cells carry oxygen around the body, and donated blood is still the only source for routine transfusions. Researchers are trying to grow red blood cells in a laboratory from stem cells. NHS Blood and Transplant says red cells can be grown from blood stem cells taken from adult blood, and it hopes to develop a new transfusion product for the small number of patients with rare blood group types that it cannot currently serve (NHS Blood and Transplant). Other laboratory work starts from induced pluripotent stem cells (iPSCs), which are adult cells reprogrammed to an embryonic-like state (Varga et al., 2025).

Established evidence

Two peer-reviewed reviews give the basic picture. Lim and colleagues (2021) state that a single unit of blood contains 2 × 10^12 enucleated red cells, and that hospitals acquire donated blood at approximately US$200 to 300 per unit (Lim et al., 2021).

Enucleation, the step in which a developing red cell ejects its nucleus, is a known bottleneck. Ju and colleagues (2024) report an enucleation rate of 4 to 10% for human iPSCs. Their review attributes that range to a single 2010 study, and other studies report different rates. They also state that it has proven difficult to manufacture these products on a large scale (Ju et al., 2024).

Demonstrated in studies

The RESTORE trial (UK). The University of Cambridge announced the trial on 7 November 2022, when two people had been given mini transfusions of lab-grown cells. The plan was for a minimum of 10 participants, each receiving two mini transfusions at least four months apart: one of standard donated red cells and one of lab-grown cells from the same donor. The lab-grown dose is around 5 to 10 ml. The announcement reports no untoward side effects in the two participants (University of Cambridge news release, 2022). NHS Blood and Transplant describes the trial as a phase I randomised crossover study in healthy volunteers. Its trial page says the trial is now closed to recruitment, and it gives no date for results (NHS Blood and Transplant trial page). The University of Bristol trial page says the aim is to measure how long the lab-grown cells last in the body, compared with the same amount of standard cells from the same donor (University of Bristol trial page). The Bristol page says the trial started in 2021, the Cambridge announcement of November 2022 reports its first two participants, and the NHS page gives no start date. The sources opened for this article do not report outcomes.

Lab-scale manufacture. Varga and colleagues (2025) report about 4.6 × 10^3 red cells per iPSC in an optimised, feeder-free culture, with a 40 to 70% enucleation rate. The work used five healthy-donor iPSC lines, although not every experiment used all five (Varga et al., 2025). The authors estimate that about 4.9 × 10^7 iPSCs would be needed to make one minitransfusion unit of 1 × 10^11 enucleated cells. This is an estimate derived from their measured yields, not a tested manufacturing run.

Mouse clearance. In the same paper, a small experiment with three mice per group found that iPSC-derived human red cells were cleared at a rate not significantly different from native human red cells, and that both were cleared entirely by 144 hours. Clearance in the first hour was higher for the iPSC-derived cells. Human red cells normally last about 120 days. The authors note a discrepancy between the typical lifespan of human red cells and their accelerated clearance in mice, which limits what the mouse model can show about survival (Varga et al., 2025).

Reported or speculative

• Company claim. On 7 May 2026, Scarlet Therapeutics, a Bristol company, published a press release on the Samedan Ltd website describing preclinical in vivo results. It says its lab-grown red cells achieved "a half-life comparable to normal donated red blood cells". The release gives no numerical half-life values, no animal numbers and no production volumes, and it does not mention a journal or peer review (Scarlet Therapeutics release, company source).
• Project announcement. Australian Red Cross Lifeblood and the University of Queensland are developing technology to make production of lab-grown red cells more efficient. The AABB, a trade body, reported the project on 29 September 2026. It states that it currently takes 3 weeks to produce roughly a teaspoon of blood, and that the project aims to increase yields while reducing production costs (AABB, 2026, trade body report). This is a project statement, not a peer-reviewed result.
• Cost estimate. Lim and colleagues (2021) report that most in vitro production processes conclude that one blood unit could carry a cost of US$15,000 or more, with some estimates reporting five times higher. This is a 2021 review estimate that cites earlier studies, not a measured price (Lim et al., 2021).
• Possible uses. Speculatively, longer-lasting manufactured cells might reduce how often some people with regular transfusion needs require donor blood, and a backup supply might help in emergency or remote care. Neither has been shown in the sources opened.

Practical limits

Cost. The only per-unit figure in the sources opened is the 2021 review estimate of US$15,000 or more, set against donated blood at about US$200 to 300 per unit. No verified current cost per unit was found.

Yield. The lab-scale yield of about 4.6 × 10^3 red cells per iPSC is far below the 2 × 10^12 enucleated cells in one blood unit. Varga and colleagues say stirred-tank bioreactors are anticipated to reduce variation and that they are pursuing this. The paper reports no bioreactor data (Varga et al., 2025).

Enough cells. Making a mini-dose already requires an estimated 4.9 × 10^7 iPSCs. Scaling to a full transfusion depends on enucleation and yield, both of which remain unresolved.

Immune matching. NHS Blood and Transplant says it cannot meet the transfusion requirements of a small number of patients with rare blood group types, and the stated aim is a new transfusion product for them in future (NHS Blood and Transplant). The sources opened do not test immune matching in people.

Time. The only production time in the sources opened is the Lifeblood project's statement of 3 weeks for roughly a teaspoon of blood. No time per full unit is reported.

Limits and open problems

• The trial uses small doses in healthy volunteers. It can show how the cells behave at that size, but not how a full transfusion would work in patients.
• The 2022 announcement reports no untoward side effects in the first two participants. No survival or safety results were found in the sources opened.
• Manufacturing cost per unit is unverified, and the lab-scale process has not been run in large bioreactors.

What to watch next

• Published RESTORE results, including how long lab-grown cells survive compared with donor cells. NHS Blood and Transplant gives no date for these.
• Peer-reviewed data on survival and safety in people, rather than mini-doses or animal models.
• Results from larger bioreactor runs, including the stirred-tank work that Varga and colleagues describe as in progress.
• Verified yields and costs from the Lifeblood project and from companies such as Scarlet Therapeutics, once their data are published.