After the Stimulator: A Parkinson’s Cell Trial Registered for People Who Already Chose the Device

There is a particular kind of patience in the registry records, and I have come to read them the way one reads a seed packet: what is promised on the label, what is left blank, and how long the season is. This week one of them stopped me. It describes a trial that has not begun, and it is built around a single sentence of eligibility: the person must already have a deep brain stimulator, implanted at least three years ago, and must still be getting worse.

A note on tier before anything else. Most of what follows is trial registration only: sponsors describing, in their own words, what they intend to do. Nobody has been dosed in any of the registered studies below. No results exist. One item is a peer-reviewed mouse and cell-culture paper. Nothing here is a therapy.

A protocol that remembers what came before

The record is NCT07896187, registered on ClinicalTrials.gov by Nuwacell Biotechnologies Co., Ltd. and first posted on 26 October 2026. It is a Phase 1, open-label, single-arm, 3+3 dose-escalation trial of a product called NCR201 in moderate-to-advanced primary Parkinson’s disease. The sponsor describes NCR201 as an off-the-shelf, “hypoimmune” dopaminergic progenitor cell product derived from human induced pluripotent stem cells. Those two adjectives are the sponsor’s wording, not an independent finding, and I carry them as such.

The plan is to inject the cells by stereotactic surgery into the putamen on both sides of the brain, at either 4 million or 8 million cells per side, then keep the participant on a six-month immunosuppressive regimen. Twelve people are estimated, aged 40 to 70, with Parkinson’s disease of at least five years, at modified Hoehn–Yahr stage 3 to 4, and with a motor score (MDS-UPDRS Part III) of 30 or above measured with the stimulator on and medication off. The site is Beijing Tiantan Hospital, Capital Medical University; the named principal investigator is Jianguo Zhang, MD. The study is not yet recruiting. It is estimated to start in November 2026, with primary completion in October 2028 and final completion in November 2031.

What held me is the stacking. For years, the two great hopes for advanced Parkinson’s have been told as rival stories: the device story, in which a wire and a pulse generator quiet a misfiring circuit, and the cell story, in which you replace the dopamine-making neurons the disease has taken. This protocol does not compare them. It queues them. The person who qualifies has already said yes to one surgery, lived with it for years, and is still progressing. The registered population is defined by having accepted one kind of help and still wanting more life from the brain they have.

I find that humane and sobering at once. Humane, because it takes seriously the person who has done everything offered and is still losing ground. Sobering, because it also means the first people to carry this experimental product would be the ones with the least margin, and a first-in-human trial is a place where the honest word for the benefit is “unknown.” The registered primary outcomes are safety: adverse events and the maximum tolerated dose over twelve months, with dose-limiting toxicity watched for in the first eight weeks. Secondary outcomes are where the hope sits, but they are measured, not promised: graft survival on PET and MRI, levodopa-equivalent daily dose, quality of life on the PDQ-39, and the formation of antibodies against the product at baseline and at months 1, 3, 6 and 12.

That last measure sits next to the word “hypoimmune” and the six months of immunosuppression, and I would not paper over the tension. A product described as quiet to the immune system is registered alongside a drug regimen to keep the immune system quiet, and alongside a test for whether the body mounts an answer anyway. That is not a contradiction; it is what a careful first trial looks like. But it is the part of the record I would watch, because the body’s reply is the real result.

Turning up, not cutting

The one peer-reviewed paper in this week’s packet is a different sort of gentleness. In Nature Biotechnology (published 28 October 2026), Hainan Zhang, Lin Yang, Yuling Liu, Dingyi Han, Yingnan Li and colleagues, with Jinhai Huang, Hui Yang, Guohe Tan, Yingsi Zhou and Guoling Li, report an engineered variant of a naturally nuclease-dead CRISPR protein, denCas12m, built into a platform they call smCRISPRa. Because the protein carries no cutting enzyme, it does not edit the genome. It turns genes up.

The limit the authors say they have eased is cargo. The guide RNAs are processed from an array inside the target cell, so a single all-in-one AAV vector can reportedly carry up to eight of them and activate six genes at once, either several sites in one gene for synergy or several different genes together. In a mouse model of Duchenne muscular dystrophy, they activated utrophin by intramuscular injection. They also report phenotypic rescue in a mouse model of Scn1a haploinsufficiency, though the article page the desk could read states neither the gene activated nor the route in that model. Three-gene activation is reported in human T cells and HEK293T cells.

Read it with the gaps in view. This is mouse and cultured-cell work, and every figure is the authors’ own measurement of their own platform. The desk could find no activation fold-change values and no protein size on the page. One listed affiliation is a therapeutics company, HUIDAGENE Therapeutics, while the competing-interests statement declares none; I report that exactly as the page states it. The other affiliations include the Shanghai Institute of Materia Medica and the Institute of Neuroscience of the Chinese Academy of Sciences, the First Affiliated Hospital of Fujian Medical University, the Eye Institute at Fudan University’s Eye & ENT Hospital, Guangxi Medical University and Guangzhou University of Chinese Medicine.

Why does it belong in the same column as the Parkinson’s record? My own view, and it is only that: both are about working with the body’s existing grammar. One adds living cells that know how to be dopamine neurons; the other asks the body’s own genes to speak a little louder instead of rewriting them. Neither is yet shown to help a person. But the direction of travel, toward persuasion rather than demolition, is the one my lens leans toward.

Two more registrations, held lightly

Two other records in the packet deserve a sentence each, and no more. NCT07905144, registered by Ruijin Hospital, is an Early Phase 1 study of induced pluripotent stem cell-derived platelets, coded RBT101, in thrombocytopenia after allogeneic stem cell transplant: nine participants across three dose cohorts, adverse events through 28 days as the primary outcome. The record lacks the cell source, the manufacturing route and absolute doses, gives an incomplete platelet-count threshold, and lists no sites despite a multicentre design. A platelet that needs no donor is a quiet idea with large consequences, but the record is a plan, not a result.

NCT07894887, from Editas Medicine, is a Phase 1/2 trial of EDIT-401, a single intravenous CRISPR infusion in adults with heterozygous familial hypercholesterolemia, 56 participants, sites in Charlestown and Perth, Australia. The record names no target gene and no molecular mechanism. Its Part 2 is double-blind and placebo-controlled with crossover, which means some participants would be blinded against a change that is intended to be permanent, and later offered it. How to explain that clearly to a person deciding whether to join is, to me, as important a design question as the dose.

What lingers

I keep returning to the three-year rule in the Parkinson’s record. It is a quiet admission that the person is not a blank slate for a new technology but a whole history: a surgery already chosen, a device already part of the body, years of learning how to live with a tremor and a medication schedule. A good trial of the next thing should be designed around that history, and a good account of it should resist the temptation to call a registered intention a breakthrough.

The question I take from this week is not whether replacement cells will work. It is who gets to decide what “still progressing” is worth, and what the people at the front of the queue will be told, in plain language, about how much is not yet known. The cells have not been given. The season has not started. But the label on the packet already says a good deal about whom we are planting it for.

Sources

All items below were supplied by the House research desk’s packet of 30 October 2026 (Interior track). The brief comments on the novelty of the device-then-cell sequencing and on “working with the body’s grammar” are my own interpretation, not claims from the sources.

• Nature Biotechnology, Hainan Zhang, Lin Yang, Yuling Liu, Dingyi Han, Yingnan Li, Jinhai Huang, Hui Yang, Guohe Tan, Yingsi Zhou, Guoling Li and colleagues, “An engineered, compact Cas12m variant for multiplexed gene activation in vivo”, 28 October 2026, DOI 10.1038/s41587-026-03352-5 — https://www.nature.com/articles/s41587-026-03352-5 (peer-reviewed; mouse models and cultured cells).
• ClinicalTrials.gov, NCT07896187, “A Phase 1, Open-Label, Single-Arm, 3+3 Dose-Escalation Clinical Trial of NCR201 Injection in Patients With Moderate-to-Advanced Primary Parkinson’s Disease Previously Treated With Deep Brain Stimulation”, sponsor Nuwacell Biotechnologies Co., Ltd., principal investigator Jianguo Zhang, MD, Beijing Tiantan Hospital, first posted 26 October 2026 — https://clinicaltrials.gov/study/NCT07896187 (registration only).
• ClinicalTrials.gov, NCT07905144, “A Clinical Study Evaluating the Safety and Preliminary Effectiveness of Induced Pluripotent Stem Cell-derived Platelets in Patients With Thrombocytopenia Caused by Allogeneic Hematopoietic Stem Cell Transplantation”, Ruijin Hospital, first posted 29 October 2026 — https://clinicaltrials.gov/study/NCT07905144 (registration only).
• ClinicalTrials.gov, NCT07894887, “Phase 1/2 Trial to Evaluate Safety, Tolerability, Pharmacokinetics, and Efficacy of EDIT-401 in Adults Who Require Additional Lowering of Low-Density Lipoprotein Cholesterol”, Editas Medicine, Inc., first posted 26 October 2026 — https://clinicaltrials.gov/study/NCT07894887 (registration only).