Seventeen babies had undergone an experimental infusion of mitochondria as of April, and 80% of them made it off ECMO — results that arrive just as November 2026 reviews are urging caution about what the technique has actually proven.
The procedure, known as mitochondrial transplantation, remains highly experimental, and its route to regulatory approval is still unclear. Yet the idea — when a battery dies, you replace it — is being tested across heart surgery, organ transplants, stroke and genetic disease. On November 2, 2026, Cheng and colleagues published a peer-reviewed review in Frontiers in Physiology focused on kidney research, distinguishing short-term biological effects from long-term therapeutic benefits that remain undemonstrated.
The stakes are anatomical as much as scientific. Roughly 10,000 infants in the United States arrive each year with severe cardiac defects that can necessitate surgical intervention shortly after delivery. Even when the operations go well, blood flowing back into the repaired heart can trigger ischemia-reperfusion injury, which damages mitochondria and ultimately kills cells.
Why mitochondria carry such weight comes down to scale.

A discovery nobody was hunting
The field’s founding surprise came in 2006, when researchers reported that mitochondria can leave their host cell and survive the journey into a different one — a phenomenon called mitochondrial transfer. The serendipity ran through James McCully’s lab at Boston Children’s Hospital, where his team found that heart-resilience work kept pointing elsewhere.
“didn't care about the mitochondria one iota,” McCully said, but “what we found was that everything we did started coming back to the mitochondria,” he said of the work with collaborator teams at Boston Children’s.
“But you know, the heart comes back — you can actually almost see it coming back” in real time, said James McCully, a research associate at Boston Children's Hospital and associate professor of surgery at Harvard Medical School, told Live Science.
The pivot came mid-experiment, when a pig heart would not restart. “And I said, 'Go upstairs; grab some of those mitochondria that we've isolated from the non-ischemic type. Just bring them down,'” McCully recalled. The heart changed color, returned to pink, and kept going; the formal rabbit-heart test was published in 2008.
That set the stage for the clinical work. The first human procedure took place in 2014, performed with Sitaram Emani, a heart surgeon at Boston Children’s, using mitochondria isolated from each patient’s own skeletal muscle.
From a rabbit heart to a Frontiers review
The developments:
- 2006 — Researchers report that mitochondria can move between cells, opening the possibility of using the transfer therapeutically.
- 2008 — The technique is formally tested in rabbit hearts.
- 2014 — The first mitochondrial transplant procedure in a baby is performed at Boston Children’s Hospital by Emani’s surgical team.
- 2023 — Treated pig kidneys, reperfused after warm ischemia, produce fewer free radicals and accrue less damage than untreated ones, reported in Annals of Surgery by Rossi and colleagues using four pigs.
- 2024 — In lab mice with a severe form of Leigh syndrome, the approach improves function and extends median lifespan by 19% to 24%.
- November 2, 2026 — Cheng and colleagues’ review in Frontiers in Physiology concludes that short-term benefits recur in preclinical kidney models, but delivery to the right renal cells remains poorly understood.
Transplant recipients weaned off ECMO successfully
The strongest human signal is retrospective. Among patients treated at a single institution between May 2002 and December 2018 who needed ECMO for severe cardiogenic shock, 8 of 10 transplant recipients (80%) weaned off successfully, against 4 of 14 controls (about 29%) — a statistically significant difference, but an association from medical records, not a randomized trial, and successful weaning is not overall survival.
Cheng and Y.-F. Liu, puts concrete numbers on the short-term promise: in a randomized swine ischemia-reperfusion study, renal-arterial injection produced a serum creatinine of 2.7 ± 0.4 mg/dL at 24 hours versus 5.6 ± 0.7 in the vehicle group (P < 0.001). But the review sets three conditions for the therapy to plausibly work — a reversible mitochondrial lesion, surviving recipient cells, and a delivery route that physically reaches them — and finds the third is the weak link, since organ-level tracking of where the mitochondria actually go has not been done.
In genetic disease, the evidence is exploratory: a 2022 study in Science Translational Medicine by Jacoby and colleagues reported compassionate use in six children with single large-scale mtDNA deletion syndrome. Separately, Giuseppe Orlando and colleagues have been cleared to run a clinical trial of mitochondrial transplantation in donor kidneys, but need more funds to proceed.
The gap between promising and proven
The Cheng and colleagues review says the field must verify separately whether mitochondria reach the target organ, which cells take them up, and what happens afterward — plus keep the administered preparation consistent and confirm long-term safety.
Lance Becker, professor and Chair of the Department of Emergency Medicine at Northwell Health and the Feinstein Institutes for Medical Research, drew a parallel with the dawn of organ transplantation: the field finds itself where medicine stood when transplanting a kidney into a patient abruptly proved transformative, and he anticipates mitochondria will trace a comparable trajectory.