Cancer diagnostics often faces challenges, such as invasiveness, high costs, and limited sensitivity for early detection, emphasizing the need for improved approaches. We present a surface-enhanced Raman scattering (SERS)-based platform leveraging inverted pyramid SU-8 nanostructured substrates fabricated via nanoimprint lithography. These substrates, characterized by sharp apices and edges, are further functionalized with (3-aminopropyl)triethoxysilane (APTES), enabling the uniform self-assembly of AuNPs to create a highly favorable configuration for enhanced SERS analysis. Performance testing of the substrates using malachite green (MG) as a model analyte demonstrated excellent detection capabilities, achieving a limit of detection as low as 10 M. Building on these results, the SERS platform was adapted for the sensitive and specific detection of hyaluronic acid (HA), a key biomarker associated with inflammation and cancer progression. The system employs a sandwich immunoassay configuration, with substrates functionalized with antibodies to capture HA molecules and 4-MBA-labeled SERS tags for detection. This setup achieved an ultra-sensitive detection limit of 10 g/mL for HA. Comprehensive characterization confirmed the uniformity and reproducibility of the SERS substrates, while validation in complex biological matrices demonstrated their robustness and reliability, highlighting their potential in cancer diagnostics and biomarker detection.