An AI algorithm, trained on thousands of samples, can analyze a simple photo of an unstained tumor slide and predict its ER-positive or ER-negative status with high confidence. This technology could revolutionize diagnostics and guide endocrine therapy in resource-limited settings where standard IHC testing is unavailable.

Trials like TaylorX and MINDACT use genomic scores to identify patients with early-stage, HR+/HER2- breast cancer who won't benefit from adjuvant chemotherapy. This avoids significant toxicity for two-thirds to over 80% of patients who would have received it under older guidelines, without compromising their outcomes.

AI identified circulating tumor DNA (ctDNA) testing as a highly sensitive method for detecting cancer recurrence earlier than scans or symptoms. Despite skepticism from oncologists who deemed it unproven, the speaker plans to use it for proactive monitoring—a strategy he would not have known about otherwise.

The Oncotype DX score effectively predicts the overall risk of recurrence for early-stage breast cancer, but it provides no information about the biological behavior of the tumor if it does recur. A tumor with a low-risk score can unfortunately return as a highly aggressive, dangerous disease, highlighting a critical limitation of the prognostic test.

Individual early-detection tests like blood biopsies or MRIs are imperfect, leading to false positives and negatives. The next step in diagnostics is a "multimodal" approach, layering different screening types, such as genomic blood tests and imaging, to create a more accurate and comprehensive picture of a patient's health.

The next frontier in preclinical research involves feeding multi-omics and spatial data from complex 3D cell models into AI algorithms. This synergy will enable a crucial shift from merely observing biological phenomena to accurately predicting therapeutic outcomes and patient responses.

The RSClin tool integrates a patient's Oncotype DX score with their unique clinical-pathologic features, such as tumor size and grade. This provides a more accurate and personalized risk assessment, as the same genomic score can represent significantly different prognoses for patients who have low versus high clinical risk factors.

The low-hanging fruit of finding a single predictive biomarker is gone. The next frontier for bioinformatics is developing complex, 'multimodal models' that integrate several data points to predict outcomes. The key challenge is creating sophisticated models that still yield practical, broadly applicable clinical insights.

While doctors focused on the immediate, successful treatment, the speaker used AI to research and plan for the low-probability but high-impact event of a cancer relapse. This involved proactively identifying advanced diagnostics (ctDNA) and compiling a list of relevant clinical trials to act on immediately if needed.