AACR Annual Meeting 2023

Precision medicine is making progress today with the convergence of richer data sets, next-generation sequencing and modeling technologies, and advanced AI/ ML methodologies. To fully realize this potential, precision medicine needs 1) “smart” diagnostics, 2) tailored therapies, and 3) appropriate timing of interventions. Each of these is challenging, but developing effective therapeutics faces the specific hurdles of increasing complexity, competition, and costs. Tempus is focused on leveraging its capabilities to help biopharma partners bring personalized medicines to patients as fast as possible in this complex environment -- in this conversation, we discuss how tools like multimodal real world data and advanced machine learning/ AI methodologies are powering next-generation drug development and driving precision medicine forward.

PARP inhibitors have demonstrated efficacy in treating solid tumors with Homologous Recombination Deficiency, the inability to repair DNA double-stranded breaks through the Homologous Recombination Repair (HRR) pathway. While BRCA1/2 are instrumental to HRR, multiple genes, including PALB2, impact the HRR pathway. No clinically approved therapies specifically targeting PALB2 currently exist, however emerging evidence suggests that patients with germline or somatic PALB2 mutations may benefit from PARP inhibitor treatment. Tempus molecular data tracking (integrated NGS and EMR data) and the TIME Trial program (rapid match of patients to Just in TIME sites for clinical trials), enabled patient identification and prescreening for this trial.

Cancer tumorigenesis and progression are driven by genetic and epigenetic alterations, giving rise to transcriptional and pathway dysregulation. The Tempus team developed a machine learning platform that integrates DNA alterations and RNA expression data to measure the activation states of oncogenic signaling pathways and characterize novel genetic alterations that may cause pathway dysregulation, and tested it on Tempus’ multimodal real-world database. Assessing both the level of pathway disruption and underlying genomic drivers may provide a more comprehensive understanding of tumor biology than assessing either factor alone.

Widespread metastatic non-small cell lung cancer is difficult to treat, yet some patients with oligometastatic disease experience prolonged progression-free survival when treated with aggressive local stereotactic body radiotherapy (RT). The research team investigated if pre-radiotherapy ctDNA testing can be used to risk-stratify those with oligometastatic NSCLC and enable earlier personalized approaches when considering systemic therapy versus aggressive local RT.

While NGS is a useful tool it can’t ascribe the cellular context for a given gene’s expression or elucidate the spatial organization of tumor microenvironments. These additional features are critical to our understanding of tumor biology and are key to the development of immuno-oncology therapeutics. The research team used the CytAssist platform to generate spatial transcriptomic data from patient samples, and used this information to create super resolution spatial transcriptomic maps.

Traditional metrics used to assess in vitro susceptibility for a given therapeutic candidate such as IC50, Emax, and AUC values do not account for cell proliferation rates, which can differ substantially when comparing multiple in vitro patient derived models. The Tempus team developed a computer vision model that utilizes label-free longitudinal light microscopy to report the number of nuclei present in organoid models across multiple cancer types, allowing for the incorporation of cell division and cytostatic effects in drug response.