Can Neck Lymphatic Surgery Reverse Alzheimer’s Disease? What the Human Evidence Actually Shows

No controlled human study has established that neck lymphatic surgery reverses Alzheimer’s disease. The procedure is real, its proposed mechanism is biologically plausible, and early research contains findings worth investigating. But improvements observed after an operation are not necessarily improvements caused by that operation.

The experimental procedure is called deep cervical lymphatic-venous anastomosis (dcLVA). It connects lymphatic vessels or structures near lymph nodes in the neck to nearby veins. Researchers hope that creating an additional drainage route will help the brain clear fluid and disease-associated proteins.

Two 2026 papers are central to the debate:

• A prospective study of 139 patients with severe Alzheimer’s disease reported a 1.28-point average improvement on a 30-point cognitive test at six months. Only 95 of the original 139 patients completed that assessment, and 44 patients experienced postoperative delirium.
• A PET imaging study involving eight patients reported a 36% decrease in an amyloid-related brain imaging measure after surgery. It had no untreated comparison group, so the scans cannot establish that the operation caused the change or that dementia was reversed.

On October 8, 2026, Vox brought renewed attention to striking before-and-after videos and the researchers investigating the procedure. Those videos are compelling because some patients appear to regain conversation, recognition, and everyday abilities after years of severe impairment.

The central question is whether surgery actually changed Alzheimer’s disease, or whether uncontrolled observations are being interpreted as proof of recovery. The published numbers, brain scans, anatomical research, and regulatory records do not yet resolve that question.

Evidence reviewed through October 9, 2026.

What Is Actually Proven?

The main gap is not whether brain lymphatic drainage exists. It does. The missing evidence is whether this particular surgery meaningfully improves that drainage and produces a durable benefit that patients would not otherwise experience.

What Does Neck Lymphatic Surgery Actually Do?

The operation is performed in the neck, not inside the brain. Under general anesthesia, surgeons identify small deep cervical lymphatic vessels and nearby veins, then construct a microsurgical connection intended to redirect lymphatic fluid into the venous circulation.

In the 139-patient study, the procedure was performed on both sides of the neck. The team used ultrasound and fluorescent dye imaging to identify vessels and verify that the new connections were open during surgery. The mean operation took approximately 212 minutes, with 286 minutes of anesthesia. This is a substantial surgical and anesthetic intervention, even though it does not involve opening the skull. The original paper describes the procedure and operative measurements.

The technique draws on lymphovenous anastomosis, a type of microsurgery used in selected cases of lymphedema, where impaired lymph drainage causes tissue swelling.

There is a critical difference between the two applications. In lymphedema, a drainage problem may be identifiable in affected tissue. In Alzheimer’s disease, researchers have not demonstrated that every patient has a surgically correctable neck lymphatic obstruction, or that placing a bypass at a particular location restores the relevant flow from the brain.

An open surgical connection is therefore not proof of a useful therapeutic connection. The decisive questions are whether flow remains favorable, whether the connection stays open, whether additional brain-derived material reaches the bloodstream, and whether any change improves patients’ lives.

Why Brain Drainage Is a Credible Research Target

The brain uses several overlapping systems to manage fluid and remove metabolic byproducts. These include exchange between cerebrospinal fluid and interstitial fluid, pathways often discussed under the term glymphatic system, and lymphatic vessels within the meninges, the membranes surrounding the brain.

Some of those lymphatic pathways connect to cervical lymph nodes. Researchers have investigated whether impaired clearance contributes to the accumulation of amyloid-beta and tau proteins associated with Alzheimer’s disease.

A 2018 study in Nature found that disrupting meningeal lymphatic function in mice impaired aspects of brain clearance and aggravated amyloid-related pathology in Alzheimer’s models. Enhancing lymphatic function also improved certain physiological and behavioral outcomes in aged mice.

That is important mechanistic evidence. It is not proof that a vein-to-lymphatic connection in a human neck can restore memory or repair neuronal loss from advanced Alzheimer’s disease.

New human evidence from September 2026

Another relevant paper appeared on September 29, 2026, in Neuron: "Cervical lymphadenectomy impairs brain lymphatic clearance and synaptic proteostasis".

Its researchers examined patients who had undergone removal of cervical lymph nodes, often as part of cancer treatment, and conducted experiments to explore possible mechanisms.

The human work included:

• 1,035 patients whose records were analyzed after neck lymph-node removal without accompanying chemotherapy or radiotherapy. The investigators observed higher rates of subsequent cognitive impairment than reported population estimates.
• 59 additional patients with brain scans before and after neck lymph-node surgery. The researchers reported accelerated widening of the temporal horns of the brain’s ventricles, an imaging finding associated with tissue loss, particularly after bilateral surgery.

The team also studied the biological effects of disrupting lymphatic drainage in rodents. The experiments implicated metabolic stress, inflammation, and damage to proteins involved in neuronal communication. Weill Cornell Medicine summarizes the patient groups, findings, and study limitations.

The distinction matters: these people did not undergo an Alzheimer’s lymphatic bypass operation. Removing lymph nodes is not the biological inverse of constructing a lymphatic-to-vein bypass. The clinical observations were also nonrandomized, and comparisons with population estimates can be affected by cancer-related, surgical, and other patient differences.

The study adds evidence that cervical lymphatic pathways matter to brain health. It does not demonstrate that the proposed Alzheimer’s surgery corrects an obstruction or reverses dementia.

What the 139-Patient Alzheimer’s Study Found

The most substantial published prospective human dataset for dcLVA was published February 3, 2026, in Alzheimer’s & Dementia by Xiaohong Fu and colleagues.

The full paper reports outcomes after surgery at the Third Affiliated Hospital of Zunyi Medical University in China. The group enrolled patients during October 2024 to March 2025. Of 142 initially enrolled, 139 formed the reported surgical cohort.

All 139 were classified as having severe Alzheimer’s disease, with a Clinical Dementia Rating score of 3. Their median baseline Mini-Mental State Examination (MMSE) score was 2 out of 30, indicating profound impairment.

The study was prospective and single-arm: investigators measured patients before and after surgery, but did not randomly assign comparable patients to surgery versus usual care or a sham procedure. There was therefore no concurrent untreated group for estimating what changes would have occurred without the operation.

Cognitive improvement was small on the 30-point scale

Source: Fu et al., primary study and follow-up figures. These figures are changes in test points, not percentages of recovered cognition.

At six months, the authors reported a 95% confidence interval of approximately +0.31 to +2.34 MMSE points and a within-patient p-value of 0.016. The result met their reported test for statistical significance.

That statistical result answers a narrow question: did the assessed participants’ test scores, on average, differ from their baseline scores? It does not answer whether surgery caused the difference.

The very low starting scores also create a floor effect. When patients begin near zero, the MMSE has limited ability to distinguish degrees of profound impairment. A small increase may be meaningful in a particular patient’s daily life, but it is not equivalent to recovery of normal memory, reasoning, or independence.

The study also reported improvements in daily functioning, behavioral symptoms, caregiver distress, and sleep-related measures. These observations deserve investigation, but without an untreated comparison group they remain postoperative associations, not demonstrated treatment effects.

Nearly one-third were not assessed at six months

The study’s participant flowchart and methods attribute the losses to difficulties such as contact problems, caregiver withdrawal, travel, and financial burdens. The authors say no patient discontinued because of a surgical complication or clinical deterioration.

There is no justification for assuming those 44 patients deteriorated. But there is equally no justification for assuming they experienced the same results as the 95 who completed follow-up.

This is attrition bias: when a large share of a study is unavailable for later assessment, the remaining participants may not represent the full group. The six-month estimate must be understood in light of its incomplete follow-up.

The operation produced significant short-term disturbances

The study abstract says no deaths or serious procedure-related adverse events occurred during the six-month observation period. Its detailed adverse-event table also records the following:

Source: Fu et al., perioperative outcomes table.

The authors describe the delirium as transient, generally resolving within three to seven days, and report no lasting consequences from those episodes. They also describe the bleeding case as associated with anticoagulant management.

Those details are important. But delirium is an acute disturbance in attention and awareness, especially consequential in someone with advanced dementia. A reoperation is also a meaningful complication, even if the investigators did not classify it as a serious procedure-related adverse event.

The appropriate conclusion is not that the surgery was uniformly unsafe, or that it was proved safe. The published cohort documents technical feasibility and short-term outcomes in a monitored setting, alongside substantial postoperative morbidity and incomplete longer-term follow-up.

Do the Biomarkers Prove That Alzheimer’s Proteins Were Drained Away?

The 139-patient study also reported changes in molecules linked to Alzheimer’s disease.

At 48 hours, concentrations of selected amyloid-beta and phosphorylated tau measures were lower in cerebrospinal fluid. Some related measurements were higher in blood. Investigators interpreted this pattern as potentially consistent with increased movement of disease-associated proteins out of the central nervous system.

That interpretation is a hypothesis, not a direct measurement of accumulated brain plaques traveling through the newly created neck connection.

Changes in protein concentration can be affected by multiple physiological processes, especially around surgery and anesthesia. Further, these biomarker assessments involved much smaller groups than the 139-patient cohort: the cerebrospinal-fluid comparison at 48 hours involved 29 patients, and several six-month blood-biomarker analyses involved 22. The published figure legend gives the measurement-specific sample sizes.

The overall biomarker changes did not reliably track cognition

The investigators tested whether changes in the measured biomarkers were associated with changes in MMSE scores. Their overall analysis found no statistically significant correlations at 48 hours or six months after adjustment for multiple comparisons.

An exploratory analysis among female patients did find an association between lower six-month plasma p-tau217 and changes in cognitive and daily-function scores. No equivalent association was reported among male patients. That subgroup finding merits replication; it does not overturn the null overall result. Both findings appear in the original paper’s correlation analysis.

The authors also compared selected biomarker trends with previously published data on people treated with lecanemab. This was an external, exploratory comparison across different cohorts, not a randomized head-to-head trial or a concurrent control arm. It cannot demonstrate that surgery performs as well as, or better than, an approved drug.

In short, the biomarker signals make further study reasonable. They do not yet establish the proposed clearance mechanism, durable disease modification, or reversal of Alzheimer’s pathology.

What Did the Eight-Patient PET Brain Scan Study Show?

A separate paper by Xu Wang and colleagues, published in Nuclear Medicine and Biology in September 2026, offers the most visually striking quantitative result so far.

Its original research report combined experiments in rats with PET imaging in eight human patients with Alzheimer’s disease who underwent a cervical lymphatic-venous operation.

The researchers reported two notable before-and-after imaging changes:

The paper reports statistically significant within-patient imaging changes and correlations between imaging differences and short-term cognitive results. Its rodent experiments further support the possibility that altered lymphatic drainage affects brain physiology.

These are interesting findings. They are also exceptionally preliminary.

A 36% PET reduction does not mean 36% of plaques were physically removed

Amyloid PET uses a radioactive tracer to estimate characteristics associated with amyloid deposition in living tissue. A change in tracer signal is an imaging result. It is not direct visualization of plaques moving through a lymphatic vessel, and it does not itself determine the exact quantity or pathway of material eliminated from the brain.

The same distinction applies to the reported MeDAS signal. Increased tracer uptake does not, on its own, establish that lost myelin was regenerated or that damaged neuronal circuits were restored.

Most importantly, there was no untreated human control group. With only eight patients, investigators cannot reliably separate the procedure’s effect from imaging variability, other physiological influences, patient selection, or changes that might occur without the operation. A low p-value does not correct the lack of a counterfactual comparison.

The authors themselves call for larger studies with appropriate controls and randomized designs. The evidence supports investigation, not a claim that an Alzheimer’s brain was rapidly restored to health.

Why the Viral Before-and-After Videos Are Not Proof

A patient who appears unable to communicate before surgery and conversational afterward is an emotionally compelling observation. It may represent a genuine improvement in alertness, speech, engagement, or another function.

But a video does not establish the cause, durability, or generalizability of that change. The critical missing information includes how patients were selected, whether the recordings were made under comparable conditions, how often less dramatic outcomes occurred, and whether independent assessors documented sustained changes across the entire treated group.

Severe dementia can coexist with fluctuations in alertness and other medical or behavioral factors. Caregiver expectations and unblinded assessments can also influence subjective outcomes. None of those possibilities proves the videos are staged or the reported improvements imaginary. They explain why video evidence cannot substitute for a controlled trial.

A spectacular individual case and an average +1.28 MMSE points among six-month completers can both appear in the same research program. The first does not cancel out the second.

Why Did China Ban Neck Lymphatic Surgery for Alzheimer’s?

China’s National Health Commission issued its restriction on June 28, 2025, and published it on July 8, 2025. The official notice prohibits using deep cervical lymphatic vessel/node-to-vein anastomosis as a clinical treatment for Alzheimer’s disease.

The government’s stated reasons were specific: the technique was still in an early exploratory phase; indications and contraindications were unclear; and high-quality evidence of safety and effectiveness was insufficient. Its official explanatory document also raised concerns about the lack of adequate supporting preclinical and health-economic evidence.

The notice does not say that the operation was proved fraudulent. Nor does it prohibit all research. It directs institutions to stop routine clinical application while describing a conditional route for scientifically justified, appropriately designed studies subject to rigorous ethics review and sufficient supporting preclinical evidence.

That is a distinction between using an unproven procedure as treatment and investigating whether the procedure can become a treatment.

The regulatory chronology resolves an apparent contradiction

The timeline matters because the existence of a 2026 publication does not mean that surgery was newly authorized for routine use in China. A paper’s publication date and the dates on which its patients underwent procedures are different facts.

Is the Surgery Approved or Being Tested in the United States?

It is being studied; it is not an FDA-approved Alzheimer’s treatment.

The relevant U.S. registry entry is ClinicalTrials.gov study NCT07178210, "Robotic-Enabled Microsurgical Intervention for Neurodegenerative Disease".

As listed in its August 2026 update, the study’s defining features include:

The distinction in the patient population is especially important. The large Chinese report involved severe Alzheimer’s disease, while the registered U.S. study targets mild to moderate disease with evidence of relevant lymphatic abnormalities. Their results cannot automatically be generalized across those groups.

According to Vox’s October 8 reporting, the U.S. investigators have performed initial procedures and hope to progress to larger, more rigorous studies if the feasibility findings support that next step.

The study also involves a company with a commercial interest in the robotic technology. That relationship is not evidence of misconduct; it is relevant context when interpreting company announcements about investigational approvals and future applications.

An Investigational Device Exemption (IDE) permits a device to be evaluated under a clinical research protocol. The FDA’s explanation of the IDE system makes clear that such studies gather information needed to assess safety and effectiveness. Authorization to conduct research is not a finding that the operation is safe, effective, or approved for general clinical treatment.

A 15-person study focused primarily on short-term device safety may answer whether the procedure is feasible to study further. By itself, it cannot establish reversal of Alzheimer’s disease.

The Key Biological Questions Are Still Unanswered

Several questions must be resolved before the operation’s proposed mechanism can be accepted as a treatment mechanism:

1. Where is the actual bottleneck? Researchers must demonstrate that the patients being treated have an abnormal drainage pathway and identify where a bypass could improve it.
2. Does the pressure gradient favor drainage? A lymphatic-to-vein connection does not automatically create useful one-way flow under varying venous pressures.
3. Does the connection remain open and functional? Intraoperative dye flow does not establish long-term patency or sustained brain-derived clearance.
4. Can the surgery change pathology that is already advanced? Alzheimer’s involves amyloid, tau, inflammation, neuronal dysfunction, and tissue loss. Improving one clearance pathway would not necessarily reverse damage already sustained.
5. Do imaging and biomarker changes predict patient benefit? The studies must show that biological measurements correspond to better cognition, function, and quality of life over clinically meaningful periods.

A 2026 critical review in Frontiers in Aging identifies pressure gradients, reflux risk, lymphatic contractility, long-term patency, and biomarker validation among the central uncertainties. These are not minor technicalities: they determine whether a surgically created route achieves what the treatment theory requires.

What Evidence Would Show That It Really Works?

A persuasive next-stage study would need a clearly defined group of patients, objective evidence of lymphatic dysfunction where relevant, and a comparison group that makes causal interpretation possible. Random assignment to surgery or an ethically acceptable control, blinded outcome evaluation, prespecified endpoints, full reporting of adverse events, and rigorous follow-up would substantially strengthen confidence.

Imaging should directly investigate lymphatic flow and, separately, assess amyloid and other disease-related changes. Cognitive tests should be paired with measures of everyday function that matter to patients and caregivers. Outcomes should be tracked long enough to distinguish transient postoperative changes from durable improvement or slowed decline.

Sham-controlled operations pose genuine ethical challenges, especially in vulnerable patients with dementia. If sham surgery is not appropriate, researchers still need credible randomized or otherwise well-controlled designs rather than relying on dramatic before-and-after comparisons alone.

It is also essential to distinguish three very different claims: temporary symptom improvement, slower disease progression, and reversal of established Alzheimer’s disease. Proving one would not automatically prove the others.

Should Patients or Families Seek This Procedure Now?

The available evidence does not justify presenting dcLVA as an established Alzheimer’s treatment or paying for it on the promise that lost memory will return. A clinician offering the procedure should be able to explain its investigational status, the evidence supporting the proposed mechanism, the relevant surgical and anesthetic risks, how outcomes will be independently assessed, and what ethics oversight applies.

Families considering research participation should discuss the decision with the patient’s neurologist or dementia specialist and consult the official clinical study record for eligibility and current status. A trial listing is not a guarantee of benefit, and research participation can involve risks.

Established Alzheimer’s care is not the same as this experimental operation. Certain FDA-approved therapies, including lecanemab and donanemab, have demonstrated slowing of decline in selected patients with early Alzheimer’s disease in controlled trials; they have not been shown to restore lost cognition or cure Alzheimer’s disease. Their suitability and risks require individualized medical assessment. See the FDA’s description of donanemab trial evidence and the Alzheimer’s Association’s treatment overview.

The appropriate response to the neck surgery findings is continued research with strong safeguards, not certainty that an apparent breakthrough has already arrived.

Bottom Line: Promising Biology Is Not Proof of Reversal

The proposed surgery rests on real anatomical research. The human studies have produced potentially important observations. Dismissing every finding because the operation is controversial would be as unscientific as accepting its most dramatic claims because the videos are persuasive.

But the evidence currently supports a narrow conclusion: some measured outcomes changed after surgery, and researchers have not yet established what portion of those changes the operation caused.

The 139-patient study found a modest average cognitive difference among participants who remained under observation, with nearly one-third missing at six months and substantial early postoperative disturbances. The eight-patient PET study found striking imaging changes but no untreated control. China’s treatment restriction and the U.S. early feasibility study both reflect the same unresolved distinction between a testable hypothesis and an established therapy.

Neck lymphatic surgery may eventually become relevant to Alzheimer’s treatment. As of October 9, 2026, it has not been shown to reverse the disease.

References and Further Reading

Original Human Studies and Biological Research

• Fu et al., "Deep cervical lymphatic-venous anastomosis attenuates cognitive dysfunction and biomarker abnormalities in severe Alzheimer’s disease: A prospective single-arm study" (Alzheimer’s & Dementia, February 2026). The original 139-patient research report, including cognitive results, follow-up counts, biomarker subsets, surgical details, and adverse events. An open-access full-text version is also available.
• Wang et al., "Translational imaging evidence for rapid reversal of amyloid and myelin pathology following deep cervical lymphatic-venous anastomosis in Alzheimer’s disease" (Nuclear Medicine and Biology, September 2026). The peer-reviewed eight-patient PET imaging study and associated rodent experiments. The publicly accessible abstract establishes the reported imaging figures; it does not provide a controlled human efficacy result.
• Khan et al., "Cervical lymphadenectomy impairs brain lymphatic clearance and synaptic proteostasis" (Neuron, September 29, 2026). Original animal and observational human evidence about consequences of cervical lymph-node removal, not a test of the Alzheimer’s bypass surgery.
• Da Mesquita et al., "Functional aspects of meningeal lymphatics in ageing and Alzheimer’s disease" (Nature, 2018). Foundational experimental research on meningeal lymphatic function, brain clearance, and amyloid-related pathology in mice.

Official Policy and Clinical Trial Records

• China National Health Commission, notice prohibiting the clinical application of deep cervical lymphatic vessel/node-to-vein anastomosis for Alzheimer’s treatment (published July 8, 2025). The primary regulatory text, including the conditions under which research may be considered.
• China National Health Commission, explanatory interpretation of the Alzheimer’s surgery restriction (July 8, 2025). Explains the agency’s concerns about safety, effectiveness, indications, preclinical evidence, and clinical use.
• ClinicalTrials.gov, NCT07178210: "Robotic-Enabled Microsurgical Intervention for Neurodegenerative Disease". U.S. investigational study registration; lists the sponsor, planned enrollment, eligibility criteria, participating sites, study design, and outcomes.
• U.S. Food and Drug Administration, "Investigational Device Exemption (IDE)". Defines the regulatory distinction between permission to study an investigational device and authorization to market it as an established treatment.

Independent Reporting and Expert Context

• Vox, "Is this the cure for Alzheimer’s?" (October 8, 2026). Original reporting on viral patient videos, international scientific skepticism, and the U.S. investigational research program.
• Weill Cornell Medicine, "Neck Lymph Node Removal Linked to Increased Risk of Cognitive Decline" (September 29, 2026). Institution’s account of the observational cohorts and animal experiments underlying the new Neuron study.
• Li et al., "Deep cervical lymphaticovenous anastomosis for Alzheimer’s disease: theoretical foundations, regulatory suspension, and translational challenges" (Frontiers in Aging, June 2026). Critical review of the mechanistic uncertainties, evidence limitations, and questions that prospective clinical research must resolve.
• Alzheimer’s Association, "Treatments for Alzheimer’s". Context on established care and the distinction between slowing disease progression and curing or reversing it.

Editorial currency note: This article reflects studies and regulatory information available through October 9, 2026. Clinical study enrollment, published results, and policy status can change. Readers should consult the linked original records for subsequent updates.