Dipanjan Pan, PhD, on early-stage tear biosensor research, closed-loop drug delivery, and the road to FDA clearance

COSMIC biosensing platform design and electrochemical sensing approach

Tears have long been overlooked as a diagnostic medium, yet their potential to reflect systemic disease continuously and non-invasively is now drawing serious federal investment. Ophthalmology Times spoke with Dipanjan Pan, PhD, FRSC, FAIMBE, FAHA, FACC Dorothy Foehr Huck and J. Lloyd Huck Chair Professor in Nanomedicine at Pennsylvania State University and professor of nuclear engineering, chemistry, biomedical engineering, and materials science and engineering, whose lab is leading biosensor development for the ARPA-H-funded COSMIC team.

Pan's group brings prior experience detecting biomolecules in tears to the OCULAB program, a $75.8 million ARPA-H initiative aimed at developing closed-loop tear-duct biosensor systems capable of continuous disease monitoring and real-time drug delivery. In the following Q&A, Pan discusses the electrochemical sensing architecture underlying the COSMIC platform, how the system determines when to deliver medication, and what the regulatory path looks like for a device combining sensing, software, and drug delivery in a single lacrimal duct-resident unit.¹

What type of electrochemical sensing are the nanostructured electrodes using, and how is the device designed to function reliably over extended wear?

Pan: Electrochemical sensing lets us detect important biomarkers by measuring tiny electrical changes. We chose it because it is fast, sensitive, low-power, and can be built into very small devices for continuous monitoring. The most common example of electrochemical sensing is an at home glucose meter. In OCULAB COSMIC, we'll be using a similar approach adapted for continuous monitoring by designing advanced sensor surfaces that reduce biological buildup and improve long-term reliability, enabling sensitive measurements over extended periods of time. We aim for the device to remain in place for weeks to months. To help the body tolerate it, we're using biocompatible materials and advanced surface coatings designed to reduce inflammation, biological buildup, and irritation, building on the safety experience of punctal plugs already used in eye care

Which specific tear biomarkers are you measuring, and how reliable are they as clinical signals?

Pan: Did you know that tears are primarily derived from blood plasma? Subsequently, tears carry many of the same biological clues found elsewhere in the body, but in very small amounts. We are looking at panels of markers related to eye inflammation, hydration balance, metabolism, and hormonal changes. The key here is validation and reliably measure the changes at a diseased state. We will compare sensor readings against established clinical tests to determine which markers are truly reliable for patient monitoring.

The system automatically delivers medication when it detects a change in tear chemistry, but how do you make sure that threshold is clinically meaningful and not over- or under-treating patients?

Pan: We are not treating every small change as a reason to deliver medication. The system will be designed to look for clinically meaningful patterns, validated against standard care, with built-in safety limits to reduce the risk of over- or under-treatment.

As AI takes on more of the real-time treatment decision-making in devices like this, where do you draw the line on how much autonomy the system should have before a physician needs to step back in?

Pan: This is a great question. I view AI as a decision-support tool, not a replacement for physicians. AI can be very helpful in identifying patterns, predicting risk, and recommending actions, but the clinical rules, treatment limits, and safety boundaries should always be established by healthcare professionals. For a futuristic system like OCULAB, the device may eventually be trusted to make small, routine adjustments within predefined safety limits, much like an automated insulin pump. However, significant changes in treatment, unusual readings, or signs that a patient's condition is worsening should always trigger physician review. The goal is to provide clinicians with better information and help patients receive care more quickly and consistently.

The device falls into a complicated FDA category called a combination product—which typically means a longer, more complex approval process. Where are you in conversations with the FDA, and what does the realistic timeline to clinical use actually look like?

Pan: You are right. Combination products do have a more complex regulatory pathway because they combine sensing, software, and drug delivery in a single platform. That is why we are engaging regulatory experts early and designing the technology with FDA requirements in mind from the beginning. Our immediate focus is demonstrating safety, reliability, and clinical value through large animal and early clinical studies. While the exact timeline will depend on study outcomes and regulatory feedback, bringing a technology like this to routine clinical use is typically measured in years, not months. The ARPA-H program is helping accelerate development by addressing key technical and translational challenges early in the process.

REFERENCES:
1. Liu E. Materials lab in national program to develop wearable eye health system. Penn State News. Published June 8, 2026. Accessed June 15, 2026. https://www.psu.edu/news/engineering/story/materials-lab-national-program-develop-wearable-eye-health-system