Caris Life Sciences has published a new study in JAMA Network Open demonstrating the effectiveness of its artificial intelligence-based GPSai algorithm in improving the accuracy of lung cancer diagnosis.

The study, titled ‘An AI Approach to Differentiating Lung Squamous Cell Carcinoma From Metastases of Other Origins,’ highlights how integrating AI into routine molecular profiling can help identify clinically significant misdiagnoses. Researchers analysed 3,958 lung cancer cases initially diagnosed as squamous cell carcinoma (SCC) and found that 123 cases were actually metastases originating from other primary cancers, including skin, urothelial, head and neck and thymic tumours.

The findings underscore the growing role of precision medicine in oncology, as reclassification of cancer type can significantly influence treatment strategies. Among the reclassified cases, 88 patients—representing 71.5 percent—would have received different first-line treatment recommendations, potentially leading to improved clinical outcomes.

Lung squamous cell carcinoma accounts for approximately 21 percent of all lung cancer cases in the United States, according to public health data. Based on the study’s findings, researchers estimate that around 1,000 cases annually could be misdiagnosed, highlighting the scale of the issue.

Caris noted that its GPSai algorithm has already contributed to correcting thousands of cancer diagnoses since early 2024. The tool integrates AI-driven tissue-of-origin prediction with comprehensive molecular profiling and traditional pathology to provide clinicians with deeper diagnostic insights.

The company’s AI platform is embedded within its molecular profiling services, including MI Cancer Seek and MI Tumour Seek, enhancing diagnostic precision without additional cost to providers. MI Cancer Seek received approval from the US Food and Drug Administration in November 2024 and is designed to deliver comprehensive tumour profiling using advanced sequencing technologies.

The study reinforces the potential of AI in transforming cancer diagnostics, helping clinicians make more informed decisions and ensuring patients receive the most appropriate and targeted treatments.