Intracochleaire biosensing voor het in situ monitoren van elektrode-geïnduceerde ontsteking

Cochlear implants (CIs) have revolutionized hearing restoration, with over a million devices implanted globally by 2022. Despite their success, the implantation process remains invasive and is associated with insertion microtrauma, triggering inflammation and a foreign body response (FBR). This causes fibrous tissue and new bone to form around the implant, increasing electrode impedance, reducing stimulation efficiency, and compromising battery life. Trauma and inflammation can also lead to residual hearing loss, key for speech perception, and in severe cases, device extrusion. Reducing the FBR is essential to improve outcomes and enable future therapies such as regenerative treatments or advanced implants. Steroids can help manage inflammation, but long-term use is not ideal. A better approach is antifouling surface coatings that prevent protein and cell adhesion, shielding the implant from immune recognition. Hydrophilic hydrogel coatings show promise, but are often thick and risk impairing electrode function. We developed an ultrathin (≈ 6 nm) antifouling coating made of polydopamine, polyethyleneimine, and sulfobetaine methacrylate (SBMA). It significantly reduced protein adhesion and completely inhibited fibroblast attachment on platinum and silicone, without affecting electrode impedance or stimulation capacity. Applied to real CIs, it also minimized impedance increases associated with fibroblast encapsulation. Though long-term in vivo testing was not possible, the SBMA coatings in vitro performance suggests strong potential for reducing FBR without affecting stimulation. Early detection of insertion trauma is also crucial. Existing tools, such as imaging or intraoperative impedance and electrocochleography readings, are nonspecific and often detect only severe trauma. To address this, we integrated an electrochemical hydrogen peroxide (H2O2) sensor into CI arrays. By modifying the preexisting Pt/Ir electrodes with poly-o-phenylenediamine (PoPD) membranes and our antifouling coating, and incorporating an iridium oxide (IrOx) pseudo-reference electrode, we enabled sensitive peroxide detection. A custom 4-step potential waveform allowed for accurate readings. Sensors showed low detection limits (≈ 3 μM), good selectivity against biological interferents, and remained stable during short-term use. Performance held after several days of storage, and the coating was non-cytotoxic and inhibited cell adhesion. In gerbil models, these sensors detected significantly elevated H2O2 in cochleae with induced trauma, correlating with impedance increases indicative of tissue disruption. While the IrOx reference electrode drifted slightly, switching to Ag/AgCl provided stable intraoperative measurements. Results were validated by in situ spiking experiments and cross validation with an Amplex Red assay. In a larger study (16 cochleae: 10 trauma, 6 control), trauma raised H2O2 from ≈ 38 to ≈ 59 μM. Trauma severity, assessed via contrast-enhanced μCT and a clinical trauma scale, correlated with H2O2 levels. Statistical analysis showed strong correlation with trauma volume. Using receiver operating characteristic (ROC) analysis, we found that a 42 μM H2O2 threshold resulted in 80 % sensitivity and 83 % specificity for trauma detection. We extended the platform to lactate sensing by incorporating lactate oxidase into the SBMA layer. The resulting sensor had a detection limit of ≈ 21 μM and retained 85 % sensitivity after 24 hours. However, stability decreased at body temperature, and the sensor's dynamic range was narrow for physiological lactate levels. Attempts to stabilize enzymes using PEGDA/SBMA hydrogels failed due to rapid degradation as was validated by an optical assay. To support this optical sensing, we developed a cost-effective, stable assay using in-house synthesized Prussian blue nanozymes in place of horseradish peroxidase. We also explored TNF-α detection using aptamer-based sensors functionalized with redox reporter molecules by square wave voltammetry. Only RNA aptamer- based sensors showed a response with detection limits down to 50 ng/ml, but suffered from signal loss in protein-containing solutions indicating biofouling. By introducing polyethylene glycol molecules into the sensor design and switching the readout mechanism to faradaic electrochemical impedance spectroscopy (EIS), DNA aptamer (VR11) based sensors measured TNF-α down to 0.1 ng/mL. Using non-faradaic EIS on interdigitated electrodes, we achieved similar detection limits with a more viable readout method. However, performance in complex protein solutions was not tested. Despite challenges, the VR11 aptamer showed sufficient binding affinity to detect TNF-α at relevant concentrations. Overall, this work advances cochlear implant technology by integrating biocompatible coatings with biosensing capabilities. We demonstrated ultrathin antifouling layers that maintain electrode function and introduced sensors that detect inflammation-linked markers in real time. This thesis is an important step in the development of smart cochlear implants that can monitor the cochlear environment, detect trauma, and potentially adapt therapy in response.