Jairath, Neil

ImportanceThere exists substantial heterogeneity in outcomes within T stages for patients with cutaneous squamous cell carcinoma (cSCC).ObjectiveTo determine whether a customized generative pretrained transformer model, trained on a comprehensive dataset with more than 1 trillion parameters and equipped with relevant focused context and retrieval augmented generation (RAG), could excel in aggregating and interpreting vast quantities of data to develop a novel class-based risk stratification system that outperforms the current standards.Design, Setting, and ParticipantsTo build the RAG knowledge base, a systematic review of the literature was conducted that addressed risk factors for poor outcomes in cSCC. Using the RAG-enabled generative pretrained transformer (GPT) model, we developed a novel class-based risk stratification system that assigned point values for risk factors, culminating in a GPT-based prognostication system called the artificial intelligence–derived risk score (AIRIS). The system’s performance was validated on a combined prospective and retrospective cohort of 2379 primary cSCC tumors (1996-2023) with at least 36 months of follow-up, against Brigham and Women’s Hospital (BWH) and American Joint Committee on Cancer Staging Manual, eighth edition (AJCC8) systems in stratifying risk for locoregional recurrence (LR), nodal metastasis (NM), distant metastasis (DM), and disease-specific death (DSD).Main Outcomes and MeasuresPerformance metrics evaluated included distinctiveness, homogeneity, and monotonicity, as defined by the AJCC8, as well as sensitivity, specificity, positive predictive value, negative predictive value, accuracy, the area under the receiver operating characteristic curve, and concordance.ResultsThe median age at diagnosis was 73 (IQR, 64-81) years, with 38.5% female patients and 61.5% male patients. The AIRIS prognostication system demonstrated superior sensitivity across all outcomes (LR, 49.1%; NM, 73.7%; DM, 82.5%; and DSD, 72.2%) and the highest area under the receiver operating characteristic curve values (LR, 0.69; NM, 0.81; DM, 0.85; and DSD, 0.80), indicating significantly enhanced discriminative capability compared with the BWH and AJCC8 systems. While all systems were comparably distinctive, the AIRIS prognostication system consistently demonstrated the lowest proportion of tumors exhibiting poor outcomes in low-risk categories, suggesting its improved homogeneity and monotonicity.Conclusions and RelevanceThe results of this diagnostic study suggest that the AIRIS system outperforms the existing BWH and AJCC8 prognostication systems, potentially providing a more effective tool for predicting poor outcomes in cSCC. This study illustrates the potential of large language models in refining prognostic tools, offering implications for treating patients with cancer.

Abstract Intratumoral genomic heterogeneity in glioblastoma (GBM) is a barrier to overcoming radiation (RT) resistance. To discover genotype-independent mediators of RT resistance, we correlated RT resistance with the concentration of approximately 700 metabolites across 23 GBM cell lines. Purine metabolites, especially those containing the base guanine, were most correlated with RT resistance. Similarly, increased abundance of tumor purines was associated with decreased survival in GBM patients treated with RT. This relationship is causal. Purine supplementation protected RT-sensitive GBMs from RT and promoted the repair of RT-induced double strand DNA breaks (DSBs). In vitro and in vivo stable isotope tracing confirmed that GBM cell lines and orthotopic patient-derived xenografts primarily generated purines through the de novo synthetic pathway. RT treatment further increased de novo purine synthesis in GBM through signaling via the DNA damage response. Inhibition of de novo GTP synthesis with mycophenolic acid (MPA) sensitized multiple GBM cell lines and neurospheres to RT by slowing the repair of RT-induced DSBs. MPA-induced radiosensitization was GTP-dependent as it was rescued by nucleoside supplementation. Modulating pyrimidine metabolism affected neither RT resistance nor DSB repair, suggesting these GTP-specific effects are due to active signaling rather than its ability to act as a physical substrate for DNA repair and candidate signaling molecules have been identified. These results were recapitulated in vivo with mycophenolate mofetil (MMF), the orally bioavailable FDA-approved prodrug of MPA. MMF potentiated RT efficacy, reduced tumor guanylates and slowed the repair of RT-induced DSBs across multiple models. Because de novo purine synthesis is activated by many of the oncogenic alterations that drive GBM, its inhibition is a promising genotype-independent strategy to overcome GBM RT resistance. We have now begun a clinical trial to determine whether combining MMF and RT is safe and potentially efficacious in patients with GBM.