Lunit (KRX:328130.KQ), a leading provider of AI-driven solutions for cancer diagnosis and treatment, will present seven posters at the American Association for Cancer Research (AACR) Annual Meeting, April 25-30 in Chicago, Illinois. The presentation will introduce the latest research from Lunit in AI-based Histopathology. It features research featuring a Lunit scope suite across a range of cancers. From rare salivary gland tumors to common types such as lung cancer.
The poster lineup includes two collaborative research with global pharma leaders, including AstraZeneca. One study conducted in a partnership with Astrazeneca will showcase the development and validation of AI models predicting EGFR mutations from H&E slides, enabling faster, more accessible mutation testing in NSCLC patients. Another co-authored with leading global biotechnology companies will apply Runittoscope IO to Phase II and III clinical trial data to predict the benefits of atezolizumab and reveal that AI-based histological profiling will help stratify patients based on the likelihood of immunotherapy response.
Of the seven studies presented, three have shocking findings with specific clinical and scientific relevance.
One study addresses the challenge of predicting response to neoadjuvant immunochemotherapy in patients with resectable salivary adenocarcinoma (SGC). This is a rare and aggressive cancer. To better understand treatment outcomes, this study applied a multimodal approach combining single-cell RNA sequencing using LunitScope®, T-cell receptor (TCR) analysis, spatial transcriptomics (Xenium), and AI-driven tissue profiling on surgically contracted tumor samples. Respondents were found to have more CD8+ dysfunction and memory T cells, along with increased TCR clonality and decreased diversity. In contrast, non-responders showed a higher presence of tumor-associated macrophages. The Lunit Scope IO contributed to detailed morphological profiling, allowing for cell type identification and spatial patterns to be verified. These findings suggest that combining advanced molecular tools with AI-driven histology may help tumor microenvironment predict responses to immunotherapy.
Another study explores potential biomarkers associated with treatment resistance in salivary duct carcinoma (SDC) by combining Lunit range IO with Xenium spatial transcriptomics. The researchers analyzed more than 915,000 cells from surgically resected salivary gland tumors, including cases of SDC treated with neoadjuvant immunotherapy. In one recurring case, the tumor showed higher expression of genes associated with immune evasion and epithelium-to-lobular transition (EMT). It is also observed that the presence of CXCL9-expressing tumor-inflammatory lymphocytes is low, which may reflect the immunologically active tumor microenvironment. These findings provide additional insight into potential resistance-related features that may not be apparent through traditional histology.
In the third study, Lunit developed an AI model to identify EGFR-mutant NSCLC tumors with similar morphological features as small cell lung cancer (SCLC), a pattern that is clinically related to early histological transformation from NSCLC to SCLC and resistance to EGFR tyrosine kinase inhibitor (TKIS) (TKIS). In this study, we used deep learning to analyze H&E stained tumor slides from 106 advanced stage EGFR mutant NSCLC patients, performing cell-level tumor heterogeneity analysis based on AI-discovered morphological features. The top 25% of patients with SCLC-like morphology (defined as the SCLC-like group) have significantly smaller nuclear areas (56 µm² vs. 102 µm²) and more intense nuclear staining. Clinically, they experienced a shorter progression of survival after TKI therapy and were more likely to convert to SCLC at libeopsy (15.8% vs. 2.0%). This study is the first to demonstrate that AI-based morphological profiling at diagnosis can identify patients at risk for small cell transformation and early TKI resistance, providing a new avenue for risk-adapted treatment plans.
The remaining research further demonstrates the breadth of Lunit’s research capabilities and AI expertise. These include studies on cell surface target discovery in prostate cancer and on strengthening preclinical immunotherapy in colon cancer.
“AACR 2025 shows how Lunit’s AI Technologies is driving a new wave of biomarker discovery and clinical insights,” says Brandon Suh, CEO of Lunit. “From salivary gland cancer to lung cancer, our study will uncover how important tumor microenvironmental patterns and risk of transformation through AI-driven histopathology, particularly Tsukii’s range IO, and even reveal how tumors respond to targeted therapy before they show that clinical progression plays a more accurate role.
