Brain tumor classification from FFPE samples using nanopore methylation sequencing

Oxford Nanopore Technology (ONT)-based methylation sequencing is emerging as a powerful approach for the rapid and accurate classification of brain tumors, an essential component of precision oncology. However, its broader clinical adoption has been limited by reliance on fresh-frozen (FF) tissue, whereas the vast majority of clinical specimens are formalin-fixed paraffin-embedded (FFPE). In this study, we address this limitation by evaluating the effects of FFPE processing on DNA methylation profiles and introducing a validated protocol for ONT-based classification using DNA extracted directly from pathology-marked regions on stained FFPE slides. This approach enables the targeted selection of tumor-rich areas following histological assessment, thereby improving DNA input quality and tumor content. We demonstrate that even small, low-input samples (≥25 ng) can be successfully classified using this method, with high concordance to final integrated neuropathological diagnoses. Our results show that, despite modest methylation loss associated with formalin fixation, classification performance remains robust. Notably, we identify a correlation between methylation degradation and fixation time, supporting a recommendation to limit formalin exposure to ≤3-4 days when possible. By enabling accurate methylation-based tumor classification from routinely processed, stained FFPE tissue, our protocol integrates seamlessly into existing clinical workflows. This expands the accessibility of ONT-based diagnostics and supports informed, timely treatment decisions-even in cases with minimal tissue availability or urgent clinical need.