Longevity research: what has reached human trials?

Longevitytech.fund — supporting longevity research

We are proud of Longevitytech.fund’s support for research into telomerase mRNA and SIRT6, two approaches explored in this guide. Rejuvenation Technologies named Longevity Tech Fund among participants in its September 6, 2023 seed financing. The fund’s September 14, 2022 announcement also described its investment in Genflow. These programs investigate how cells maintain tissues and respond to age-related damage. [17] [18]

The fund’s portfolio also includes Intervene Immune, Repair Biotechnologies and Klothea Bio, whose programs appear in this overview. We are proud to support these distinct paths toward preserving health and function. [23]

The research is real, but it asks several different questions. Restoring function in a damaged optic nerve, treating a severe inherited cholesterol disorder, and changing an immune-aging measurement are separate goals. A result for one cannot be assigned to the others. This guide compares nine approaches, including academic work and companies beyond the fund’s network. Evidence dates extend through September 22, 2026.

What human studies have established

• Shown: several approaches have reached human testing, and some have published results. Those results include encouraging observations, small safety studies, and unsuccessful primary comparisons. [2] [3] [6] [8] [9] [10] [14] [15]

• Open questions: depending on the program, evidence is still needed for improved vision, immune function, tissue recovery or disease prevention. Trial registration and first dosing are development milestones; patient outcomes determine clinical value. [1] [16]

• What would change the assessment: reproducible improvements in a defined patient outcome, accompanied by adequate follow-up for harms. The decisive result might be better vision or function, rather than a claim about aging as a whole.

Nine approaches and the questions still open

The dates below identify reports or trial-record updates. They are not expected treatment-launch dates.

A trial phase tells you what is being attempted

In the US drug-development framework, Phase 1 generally emphasizes safety, tolerability, and dose selection. Later studies test benefit and risk in more people. The actual design still matters: an uncontrolled study and a randomized comparison cannot support the same causal conclusion simply because both involve patients. [1]

Registration, treatment, results, and approval are also separate events. A registered trial may not yet have enrolled anyone. A study can finish before results are reported. Authorization to conduct research does not authorize marketing for a particular use. A company accepting inquiries establishes neither routine availability nor suitability for the person asking. [1]

ER-100 illustrates a clear milestone with a limited meaning: its sponsor reported that a participant received treatment. That is a meaningful step beyond animal experiments. The announcement did not provide a numerical safety dataset or show that vision improved. [2] [3]

When descriptions of clinical progress disagree

Two programs in this comparison have unresolved public-status discrepancies. For AKL003, the later registry lists a planned study at GARM in Roatán, Honduras, as not yet recruiting, whereas the earlier company release announced recruitment. The documents also differ on aspects of eligibility and placebo description. These differences prevent a reliable conclusion about actual enrollment, dosing, or individual eligibility. [12] [13]

For Rejuvenation Technologies, the Y Combinator company profile goes beyond the main website’s inactive-trial labels. Yet it supplies no trial identifier, dosing date, oversight documentation, or human outcome dataset. The appropriate conclusion is that clinical status remains unresolved. Neither description justifies claiming proven human benefit, and the inactive labels alone cannot establish that nobody has received the intervention. [4] [20]

The profile’s reference to FDA INTERACT does not close this gap. INTERACT is a US FDA process for early development advice. It does not by itself authorize a particular human trial or confer marketing approval. [21]

What published human results actually say

The senolytic bone study provides a useful test of expectations. In 60 postmenopausal women, dasatinib plus quercetin did not significantly outperform the comparison group on the prespecified marker of bone breakdown at 20 weeks. Exploratory analyses suggested different responses among women with higher levels of a cellular-senescence marker. That could inform a future trial, but it does not replace the original primary result or establish fracture prevention. [8]

Mitochondrial transplantation shows why the patient group and comparison unit matter. The small cardiac study associated treatment with recovery in seriously ill children, using retrospective controls. It could not establish the causal certainty of a randomized trial. The reproductive study involved 151 patients and 1,178 mature eggs. Eggs from each patient were allocated to mitochondrial transfer or standard fertilization procedures; the primary day-three embryo-quality result did not improve. Later choices about which embryos to transfer depended on quality and availability, making pregnancy comparisons observational. Neither study tested general rejuvenation in older adults. [9] [10]

The GRF6019 studies answer another, narrower question. In the 2020 study, 47 people with mild-to-moderate Alzheimer’s disease were randomized between two doses and followed for 24 weeks. Without a placebo group, the study could not establish benefit over no active treatment. The 2021 severe-Alzheimer’s study included 18 active-treatment participants and eight receiving placebo, but was too small to establish differences in clinical benefit reliably. Results for GRF6019 also do not transfer to whole plasma, plasma exchange, or another fraction. [14] [15]

Unfavorable results deserve their place beside promising observations. They can narrow a hypothesis without settling every possible use of a mechanism. Inadequate delivery or the wrong patient group might explain a failure, but those possibilities need testing; they do not turn a negative result into a positive one.

Biological-age changes need a clinical meaning

A biomarker is a measurable feature of biology, such as a protein level or a pattern of chemical marks on DNA, called DNA methylation. A methylation clock uses those patterns to estimate an aspect of biological age. Such measurements can show that an intervention affected biology without proving that someone feels or functions better. [16]

This distinction is particularly important for Genflow’s canine clock result and the TRIIM immune-aging study. A decrease in an age estimate is not a measured number of life-years gained. Likewise, raising alpha Klotho protein would establish an effect on protein expression, not every benefit associated with that protein in other research. [5] [6] [12] [16]

The US FDA describes a surrogate endpoint as a measure used in place of a direct clinical outcome. Its value depends on evidence connecting changes in that measure to the intended patient benefit. Association with health or risk is an important starting point; it is not enough to prove that changing the marker changes the outcome. [16]

The cited SIRT6 laboratory paper concerns genome maintenance, rather than treatment benefit in people. Its 2023 correction removed a reference to a retracted study; the authors stated no other change to the text or interpretation. [19] [22]

Safety belongs in the comparison

A small study can identify problems while still missing uncommon, delayed, or repeated-treatment harms. In the 47-person GRF6019 study, one serious infusion-related event was considered related to treatment. Calling the preparation simply “safe” would erase information that matters. [1] [14]

Safety findings also belong to the product, route, population, and observation period studied. A risk acceptable for a severe disease may not be acceptable for an otherwise healthy adult. A treatment’s intended biological precision does not remove the need to observe what happens to people. [1]

How Healthy Longevity Clinic experts evaluate the evidence

For someone trying to preserve health over the next decade, the useful question is whether a program is moving toward a benefit that matters to that person. HLC’s interpretation starts with the link between the study population and the goal: improved vision in optic-nerve disease would be important, but it would not establish improved heart or brain health in people without that disease. [1] [2]

A second distinction is the type of outcome. A clock shift in dogs, a protein increase, and an improvement in everyday function are different findings. HLC gives the reader a way to keep them separate: identify the exact product, who received it, the comparison group, the main outcome, and the harms over the observed period. That makes a meaningful clinical conversation possible without ranking unlike experiments as competing rejuvenation treatments. [5] [12] [16]

The evidence that would change this assessment is specific: a reproducible controlled result for a defined condition, with an effect large enough to matter and follow-up adequate to assess risk. For ER-100, that means human safety and visual outcomes; for senolytics, confirmation beyond an exploratory subgroup; for plasma fractions, convincing clinical benefit from the exact preparation. Progress on any one of these questions would be valuable on its own terms. [2] [8] [14] [15]

Milestones worth watching

The dated records point to several possible next events. ER-100’s registry estimates primary completion in May 2027. Genflow scheduled a presentation for October 2 at an October 1–2, 2026 summit. AKL003’s registry estimates completion in November 2026, although its enrollment status is unresolved. These are plans, not promises of published results, approval, or access. [2] [5] [12] [13]

TRIIM-X demonstrates why old estimates need care: its registry’s December 2025 completion estimate had passed without an updated status or posted results in the September 2026 evidence. A passed date cannot establish that a trial is complete. [7]

Three questions for a clinical conversation

1. Does this study concern my condition and a patient outcome I care about, or only a laboratory measure, animal result, or different disease?

2. What did the comparison group show, and what harms were observed in how many people over what period?

3. Which result would make this intervention relevant to my care, and has that result actually been reported?

Common questions

Which approach is closest to becoming a treatment?

There is no meaningful single ranking across these nine approaches. Some have published human studies; others have protocols, animal results, or conflicting clinical-stage claims. The more useful question is which exact product has convincing benefit for a defined condition. A phase number alone cannot answer it. [1]

Does a younger biological-age score mean extra years of life?

No. A clock result is a measured change in an estimate. Translating it into longer life requires separate evidence connecting treatment-induced changes with health outcomes. [16]

Could a disease treatment later help healthy people?

It could become a research question. The disease result would still need separate testing in the new population, including a new assessment of benefit and risk. The studies here do not supply that evidence. [1] [2] [11]

Where can I see the underlying studies?

The numbered references link to the papers, trial records, and dated announcements discussed here. Read the primary outcome and comparison group alongside the headline, and distinguish planned participant numbers from actual results.