Predictive Oncology (NASDAQ: POAI) stock surged during intraday trading Tuesday after the company revealed promising results from a retrospective study conducted with UPMC Magee-Womens Hospital.
Predictive Oncology (POAI) has unveiled promising findings from a retrospective study indicating that its artificial intelligence (AI) and machine learning (ML) capabilities may enhance survival outcome predictions for ovarian cancer patients compared to traditional clinical data. The company will present the study findings at the upcoming American Society of Clinical Oncology (ASCO) Annual Meeting in Chicago, IL, from May 31 to June 4, 2024.
Study Details
The study, conducted in collaboration with UPMC Magee-Womens Hospital, focused on developing ML models to accurately forecast short-term (two-year) and long-term (five-year) survival outcomes for patients with high-grade serous ovarian cancer. This particular cancer type presents a challenge due to its typically asymptomatic early stages, leading to late diagnoses and high relapse rates. Dr. Robert Edwards from Magee-Womens Hospital highlighted that nearly 80% of patients relapse within two years, with only 20% achieving long-term survival.
The research involved the analysis of patient data and tumor specimens collected between 2010 and 2016. Researchers employed whole exome and transcriptome sequencing, drug response profiles, and digital pathology profiles to train 160 multi-omic ML models. They evaluated the performance of these models using the AUROC metric, which indicates higher predictive potential with scores above 0.5.
Key Findings
Dr. Brian Christopher Orr, MD, the study lead investigator, will present the findings at the ASCO meeting. He noted that seven of the 160 models showed high accuracy for two-year survival predictions, and 13 demonstrated high accuracy for five-year survival predictions. Importantly, the study found that multi-omic input features resulted in superior predictions compared to clinical data alone, with the top-performing models significantly outperforming any single feature set.
Predictive Oncology interprets these findings as supporting further ML model development for clinical applications and unique biomarkers identification for innovative cancer therapeutics. Leveraging its extensive biobank and AI platform PEDAL, which boasts a 92% accuracy rate in predicting drug responses, the company positions itself as a frontrunner in AI-driven drug discovery initiatives.
Predictive Oncology (NASDAQ: POAI) Stock Price Action
On Tuesday, POAI stock soared 32.84%, closing at $2.09, marking a 23.37% increase for the week. The trading volume was 96,900,184 shares, significantly higher than the average daily volume of 1.74 million.
