Could a tear sample reveal Parkinson's disease before someone develops obvious symptoms? It's the kind of question that makes a compelling startup pitch. But NeuroXense is still at a pre-clinical stage, and that fact belongs at the beginning of the story.

Camila Feltrin, co-founder and CEO of the Argentine team NeuroXense, presented its tear-based Parkinson's concept in the official Web Summit PITCH semi-final on November 12, 2026. The company did not reach the final. Patient tear-sample work was described as a next step for 2027, not a completed clinical validation.

How far did NeuroXense go at Web Summit?

The pitch took place on Centre Stage during the competition's Thursday semi-final. Carevision, Zoyk and Moola advanced to the final. NeuroXense did not, although appearing in the semi-final gave the team international exposure for a research-stage idea.

That exposure can be valuable for attracting collaborators. It is distinct from regulatory approval or proof that a test works in patients.

What is the proposed Parkinson's tear test?

The concept is to look for biological signals in tears that may be associated with Parkinson's disease. A minimally invasive sample could potentially be easier to collect than more burdensome testing, if the relevant markers prove detectable and clinically meaningful.

The attraction is obvious. Parkinson's disease can be difficult to identify early, and researchers are interested in better biomarkers. But promising a diagnosis years before symptoms would require much more than identifying a laboratory signal.

A useful diagnostic test must distinguish affected from unaffected people in real patient populations, demonstrate repeatability and perform well against an appropriate clinical reference. The point at which a laboratory hypothesis becomes a patient test is reached through that evidence, not by naming the hoped-for benefit.

What stage is the work at?

NeuroXense presented a pre-clinical concept, with patient tear-sample work planned for 2027. Detecting Parkinson's years before symptoms remains a research ambition rather than a validated capability of the proposed test.

At this stage, the scientific milestones are the quality of the biological hypothesis, the reliability of the measurement method and the design of the planned patient studies.

Camila Feltrin, co-founder and CEO of NeuroXense, during the PITCH semi-final on Centre Stage. Photo by Shauna Clinton/Web Summit via Sportsfile, CC BY 4.0, cropped
Camila Feltrin, co-founder and CEO of NeuroXense, during the PITCH semi-final on Centre Stage. Photo by Shauna Clinton/Web Summit via Sportsfile, CC BY 4.0, cropped

What evidence should come next?

The next milestones would include patient cohorts, appropriate comparison groups, biomarker reproducibility and a clear test-use case. Is the ambition screening an asymptomatic population, supporting diagnosis when symptoms appear or monitoring known disease? Those are distinct medical questions.

Tear-based analysis offers a research route into neurodegenerative disease. NeuroXense is working to test whether its proposed signals can become clinically useful information.

Why tear fluid is an interesting research material

A blood test is familiar, but collecting blood requires a procedure that may be inconvenient for screening or repeated monitoring. Researchers therefore investigate other biological materials that might contain useful signals, including saliva, breath and tears. The scientific question is whether a measurable feature in that material reliably relates to the condition of interest.

NeuroXense's proposal belongs to that line of investigation. Camila Feltrin pitched a tear-based approach to detecting signals associated with Parkinson's disease at Web Summit's PITCH semi-final. The concept is appealing because a less invasive sample could potentially make some kinds of testing easier in the future.

A research hypothesis can be compelling without having reached clinical validation. Distinguishing those stages is especially important in neurodegenerative disease, where people and families may be eager for an earlier answer.

The evidence needed for earlier detection

Parkinson's disease can develop over a long period, and researchers are interested in identifying changes before the more familiar motor symptoms become obvious. But showing that a particular biological feature might relate to the disease is not the same as establishing a reliable early-detection test.

For a claim about detection years before symptoms, researchers would need evidence connecting early measurements to later independently established outcomes. That could involve prospective follow-up or other rigorous designs able to test the timeline. A laboratory hypothesis or a small preliminary dataset cannot by itself establish that an individual would have been identified years before diagnosis.

The implications of false results are serious. A false positive could cause considerable distress, particularly when the meaning of a signal is uncertain and treatment decisions are not straightforward. A false negative could provide unwarranted reassurance. A screening test must therefore be assessed not only on its ability to distinguish samples in a controlled setting but on what its results would mean for real people.

Earlier detection is the company's goal. Establishing that capability will require evidence linking its measurements to reliably determined clinical outcomes.

What has to happen before a tear test becomes useful?

The first step is to define the biological target and the measurement process. Researchers need to establish whether the feature can be detected reliably in tear samples, how much it varies between collections and whether factors unrelated to Parkinson's affect the signal.

Then comes clinical relevance. Samples from people with and without the condition would need to be compared using a suitable reference and a study design that avoids obvious sources of bias. Ideally, evaluators would consider other conditions and circumstances that might produce similar results. A test is not clinically specific merely because it detects a difference between two narrowly selected groups.

The intended use must also be defined. Is the goal to support research, identify people who may need further assessment, help distinguish conditions or monitor people already diagnosed? These uses create very different demands. Evidence appropriate for a research instrument does not automatically justify a clinical screening claim.

If a product is eventually developed for routine use, practical issues will matter too. Tear collection must be consistent and acceptable to users. Sample handling and the equipment required for analysis will affect cost and deployment. A test promoted as simple should be evaluated across the full process, not just the final laboratory measurement.

From left, Clive Nabale of Zoyk, Hyunji Lee of Sakak, Camila Feltrin of NeuroXense and Mantas Vičius of BBright backstage of Centre Stage. Photo by Sam Barnes/Web Summit via Sportsfile, CC BY 4.0, cropped
From left, Clive Nabale of Zoyk, Hyunji Lee of Sakak, Camila Feltrin of NeuroXense and Mantas Vičius of BBright backstage of Centre Stage. Photo by Sam Barnes/Web Summit via Sportsfile, CC BY 4.0, cropped

From laboratory work to patient studies

Early research programmes depend on building a case step by step. A team may start with a biological hypothesis, develop an assay, work with laboratory materials and then move toward patient samples. Each milestone can be meaningful if described honestly.

Early research moves from biological hypotheses and measurement methods toward relevant patient studies. The long-term early-detection goal and the work completed so far are separate stages in that process.

Patient sampling is planned for 2027. Those studies would help evaluate whether the proposed signal has clinical relevance and establish performance against an appropriate reference.

What would a clinical partner want to know?

A specialist considering research collaboration would begin with the proposed target and analytical method. Is there a biologically plausible rationale? Can the signal be measured consistently? How are samples collected and stored? What controls are being used? These are ordinary scientific questions that determine whether the study can generate interpretable results.

A future healthcare buyer would need additional answers. Which patient population is appropriate? How would a positive result change care? What happens if the test disagrees with other assessments? Would its use reduce delays, or would it mainly create a new group of people requiring specialist follow-up?

Commercial viability also depends on the route to use. A test that requires sophisticated laboratory processing may have a different business model from a point-of-care device. The final format, manufacturing requirements and regulatory pathway would depend on the test's intended use.

Why the semi-final still matters

Reaching the PITCH semi-final gave NeuroXense a chance to bring an early scientific concept before a global audience. That is a legitimate achievement, even though the company did not advance to the final.

The more important story will unfold in the research milestones that follow. Can the team collect and study relevant patient samples? Does the proposed signal remain informative across people with varied characteristics? Can independent investigators reproduce the findings? And, eventually, is there a clinically useful decision that a reliable test would support?

Those studies are the route from a pre-clinical concept toward a trustworthy test. For NeuroXense, planned patient tear-sample work in 2027 is a next step in evaluating its hypothesis.