Introduction

Lung cancer, particularly non-small cell lung cancer (NSCLC), remains one of the most prevalent and lethal malignancies worldwide. Erlotinib (ERL), an epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor, was approved by the U.S. Food and Drug Administration (FDA) in November 2004 for the treatment of metastatic NSCLC. At the standard therapeutic dose of 150 mg per day, ERL steady-state trough plasma concentrations in NSCLC patients exhibit marked inter-individual variability, typically ranging from approximately 0.73 μM (315.6 ng mL-1) to 10.43 μM (4480 ng mL-1), with a minimum effective threshold of around 1.16 μM (500 ng mL-1) proposed for adequate EGFR inhibition [1-3]. This wide pharmacokinetic variability, coupled with the narrow margin between therapeutic and toxic concentrations, underscores the clinical necessity for routine therapeutic drug monitoring (TDM) and an analytical platform capable of quantifying ERL across this entire concentration span [4]. Recent clinical models continue to confirm that severe inter-individual exposure and adverse outcomes, such as interstitial lung disease, are strongly driven by these variable plasma concentrations, necessitating robust patient-centric monitoring [5-7]. Despite its therapeutic efficacy, erlotinib is associated with significant adverse effects including folliculitis, diarrhoea, dry skin, and fatigue, which may necessitate dose reductions or early discontinuation [8]. Consequently, therapeutic drug monitoring (TDM) is critical for optimizing dosing, improving therapeutic outcomes, and minimizing toxicity [9-12].

Conventional methods for ERL quantification, including high-performance liquid chromatography (HPLC), liquid chromatography-mass spectrometry (LC-MS), liquid chromatography-tandem mass spectrometry (LC-MS/MS) and others are characterized by high sensitivity but require sophisticated instrumentation, trained personnel, lengthy sample preparation and hazardous solvents [13-18]. Immunoassay-based platforms, while offering some multiplexing advantages, suffer from cross-reactivity, interference from endogenous biomolecules, and lower specificity [19-21]. These limitations underscore the need for rapid, low-cost, and field-deployable analytical platforms. Electrochemical methods have emerged as powerful alternatives owing to their simplicity, cost-effectiveness, miniaturisability and inherent compatibility with point-of-care settings [22-25]. Indeed, state-of-the-art voltammetric sensor development has aggressively targeted tyrosine kinase inhibitors using composite matrices to meet clinical accuracy demands in complex biological fluids [26-29]. Among these, disposable SPEs are particularly attractive because of their affordability, portability, and suitability for mass production [29-32].

Titanium dioxide nanoparticles (TiO₂NP) are widely employed in electrochemical sensing due to their large surface area, excellent biocompatibility, chemical stability, and pronounced electrocatalytic properties [33-34]. Modern nano-interfaced designs increasingly leverage the high surface-to-volume ratio and catalytic properties of TiO₂ on SPEs to expedite heterogeneous electron transfer rates and boost trace-level sensor sensitivity [35]. TiO₂NPs have been shown to enhance electron-transfer kinetics and improve analyte adsorption at electrode surfaces, leading to improved detection performance. Among the metal oxide nanomaterials explored for electrochemical sensing, including ZnO, CeO₂ and Fe₃O₄, TiO₂ was specifically selected in this work for three key reasons. First, anatase-phase TiO₂ exhibits exceptional chemical stability across the neutral pH range employed in physiological sensing (pH 7.4), remaining insoluble and structurally intact in phosphate buffer, unlike ZnO which undergoes surface dissolution under similar conditions [36]. Second, TiO₂ exhibits well-documented electrocatalytic activity toward the oxidation of aromatic organic compounds, attributable to its abundant surface hydroxyl groups and high adsorptivity for π-electron-rich molecules such as erlotinib [37,38]. Third, TiO₂ is established as non-toxic and biocompatible, a critical requirement for sensors intended for clinical serum analysis, where cytotoxic modifier residues could compromise sample integrity and patient safety.

In the present study, we report the fabrication and characterization of a TiO₂NP-modified SPE (TiO2NP@SPE) and evaluate its application for the ultrasensitive electrochemical determination of erlotinib. The modified sensor achieved substantially higher sensitivity and an improved electroactive surface area compared to the unmodified electrode, while shifting the linear range to a range more relevant to clinical TDM concentrations, with excellent selectivity and reproducibility in human serum samples, demonstrating its suitability for TDM in clinical practice.