Lab-grown organs

Lab-grown organs are transplantable organs produced from cells rather than taken from a donor. The category is defined by ambition rather than method: it includes organs built on a scaffold, organs grown inside a host animal, and organs regenerated in place inside the patient. As of 2026 no solid vascularized organ made by any of these routes has been transplanted into a human. What has reached patients is a narrower set of engineered tissues, and understanding why the boundary falls where it does is more useful than any timeline.

§What counts

The functional distinction is between organs that are essentially sheets or clusters and organs that are perfused three-dimensional structures with an arterial inlet and a venous outlet.

The first category has clinical products. Cultured epidermal grafts, engineered skin, cultured limbal epithelium for corneal repair, and allogeneic cultured thymus tissue for children born without a thymus are all approved somewhere. Stem-cell-derived pancreatic islet clusters, which are Organoids in construction if not in name, have produced insulin independence in most participants of small type 1 diabetes trials, using standard immunosuppression — arguably the first case of a lab-grown endocrine organ doing an organ's job in a person, even though it is delivered as a suspension of cell clusters into the portal vein rather than as an anatomical pancreas.

The second category — kidney, liver, heart, lung — has nothing. Each requires on the order of ten to a hundred billion cells of several types, arranged around a branching vascular tree fine enough that no cell sits more than a couple of hundred micrometres from a capillary, with mechanical properties that survive arterial pressure and a surface that does not clot.

§Route one: build it

The constructive route seeds cells onto a scaffold and cultures the result. Purely synthetic scaffolds have worked for hollow, low-metabolic-demand structures — bladder, urethra, vaginal tissue — and not for solid organs, for the reasons set out in Tissue engineering and Organ bioprinting. The uterus falls on the far side of that boundary, being muscular and heavily perfused, which is why the only route to a working one remains a donated organ and the procedure described in Uterus transplantation.

The variant with the strongest results uses a donor organ stripped of its cells as the scaffold, retaining the vascular tree in place of trying to fabricate one. Perfusing a detergent through the vasculature of a rat heart leaves an acellular matrix that can be reseeded with cardiac cells and stimulated to beat; the same approach has been applied to rodent lung and kidney, with reseeded kidneys producing dilute urine when perfused.¹² These constructs functioned at a small percentage of native capacity for hours. Scaling the approach to human-sized organs runs into incomplete re-endothelialization, which causes thrombosis, and into the sheer number of cells required. This route is treated in detail in Decellularized scaffolds.

§Route two: grow it in an animal

Interspecies blastocyst complementation exploits developmental vacancy. If an embryo is genetically unable to form a particular organ — a condition created by knocking out a master regulator gene with CRISPR–Cas9 or an earlier editing method — and cells from another animal are injected at the blastocyst stage, the donor cells fill the empty niche and build the organ. Rat cells injected into mouse embryos lacking Pdx1 form a rat pancreas inside the mouse; islets grown this way and transplanted into diabetic mice corrected their diabetes with only transient immunosuppression.³⁴

Extending this to human cells in pigs is far harder. Human pluripotent cells contribute at extremely low rates to pig embryos, and most undergo apoptosis; early chimera experiments detected human contributions only as a vanishingly small fraction of the embryo's cells.⁵ Anti-apoptotic modification of the human cells and better matching of developmental stage have improved this, and a 2023 report described pig embryos carrying substantially humanized mesonephroi at around four weeks of gestation.⁶ Japan lifted its prohibition on gestating human-animal chimeric embryos in 2019, and other jurisdictions have taken varied positions. The oversight questions resemble those raised by Stem-cell-based embryo models: both involve entities whose moral status depends on a definitional judgement that regulators have not settled.

Three problems remain. The organ's blood vessels and connective tissue are typically host-derived even when the parenchyma is human, so a "human" kidney grown in a pig presents pig endothelium to the recipient's immune system — which is the same barrier that Xenotransplantation addresses by editing the donor animal, and which suggests the two approaches converge rather than compete. Donor cells cannot be confined to the target organ without additional engineering, and contribution to the host's brain or germ line is the ethical trigger point that most regulators focus on. And gestating a human-scale organ in a pig requires a target organ of appropriate size and a host that survives to term.

§Route three: grow it in place

The least discussed route uses the patient's own body as the bioreactor. The liver already regenerates, and the field's interest is in extending that capacity: transdifferentiating hepatocytes or biliary cells, delivering growth signals to a damaged kidney, or implanting a small mass of cells that expands in situ. Ectopic organogenesis — growing a functional secondary organ in the omentum or lymph node, where blood supply is abundant — has produced functioning ectopic liver tissue in animals.

In situ approaches sidestep vascularization and cell manufacturing at once, because the host supplies both. They cannot replace an organ that has been removed, and they depend on regenerative capacity that mammals largely lack outside the liver, a limitation discussed in Limb regeneration.

The vascular tree is the organ

Across all three routes, the recurring finding is that parenchymal cells are the easy part. Hepatocytes, cardiomyocytes and nephron progenitors can be made in quantity from Induced pluripotent stem cells. What cannot be made is the branching, endothelialized, non-thrombogenic vasculature that keeps them alive, which is why the most promising strategies borrow a vascular tree rather than build one.

§The immunology of a grown organ

A lab-grown organ is not automatically an immunologically silent one.

An autologous organ, made from the patient's own reprogrammed cells, should escape rejection, but takes months to produce, costs a great deal per patient, and inherits the patient's genome — including whatever mutation caused the organ failure, unless it is corrected by Somatic gene therapy first. Autologous manufacture is also incompatible with acute failure, where the patient has days, and a bespoke product priced per patient raises the distribution questions collected under Access and inequality.

An allogeneic organ made from a banked cell line is manufacturable and immediately available, and needs immunosuppression like any transplant. The route being pursued to avoid that is hypoimmune engineering: deleting the genes required for surface expression of class I and class II major histocompatibility complex, and overexpressing CD47 so that natural killer cells and macrophages do not attack the resulting cells for lacking MHC. The approach was established in mice and in human cells before it reached patients.⁷ A 2025 report described hypoimmune-modified donor islet cells surviving and secreting insulin in a person with type 1 diabetes who received no immunosuppression, at short follow-up. That is early single-patient evidence rather than a demonstration of durable tolerance. Whether cloaked cells remain safe over years is unresolved, since the same modifications that hide a transplant from immune surveillance would also hide a tumour arising from it.

§Outlook

The field's realistic near-term output is not organs but organ substitutes: islet products, engineered corneal and skin tissue, bioengineered vascular grafts, cultured red cells of the kind described in Artificial blood, and external assist devices built from decellularized animal organs reseeded with human cells. For the heart, the working substitute remains mechanical, as set out in Artificial heart. For kidneys and livers, the supply pressure described in Organ shortage is being addressed by machine perfusion and donation policy, and experimentally by edited pigs, whose grafts have so far all failed or been removed within months; each has still moved faster than growing organs from cells.

The question worth watching is not when a lab-grown kidney is implanted but whether any group can keep a human-scale bioengineered organ perfused and functional in a large animal for a month. That threshold has not been crossed by any route, and until it is, the difference between a grown organ and a research construct remains the difference between a circulatory system and a scaffold.

§See also

• Organ bioprinting
• Decellularized scaffolds
• Tissue engineering
• Organoids
• Xenotransplantation
• Organ shortage
• Induced pluripotent stem cells
• Artificial heart

§References