What is precision medicine?
Precision medicine is the use of a patient’s own tissue, health records, or data to make healthcare related decisions. This can be something as common as the use of a saliva sample to learn more about family genetics or as obscure as the use of skin cells to determine how a hereditary muscle disease may progress. The objective is to use patient-specific information to make the best healthcare decisions for each individual rather than using trial and error to find treatment that works.
Cancer precision medicine is used throughout the patient journey. Tools exist for diagnosis, prognosis, treatment selection, and prediction of recurrence. We believe that the area with the greatest potential for improvement is treatment selection.
In 2021, there will be 1.9M new cancer diagnoses made in the US alone (1). Of these, approximately 57% will be diagnosed at late stage, when cancer has already spread (1). These patients could be asymptomatic, unable to recognize their symptoms, or unable to get an early diagnosis due to healthcare access. Unfortunately, those diagnosed at late stage will have a 5-year survival rate of as low as 3% (depending on disease), emphasizing the need for more effective treatment once the disease has been detected.
Current precision medicine for treatment selection
Genetic testing is used to determine which mutations are present in cancer cells. This information is valuable when determining which drug to give a patient as some mutations are “actionable”, meaning there is a drug or set of drugs that are expected to work better for patients with that mutation. Genetic testing can be run using blood (liquid biopsy) or a solid tissue sample taken directly from a patient’s tumor. After retrieval, samples are sent to a lab for sequencing and analysis. Over 300 genes are analyzed and mutation status is reported alongside treatments for actionable mutations (2).
As of today, it is believed that there are over 3,400 cancer driving mutations (3). Unfortunately, most of these mutations are not actionable and therefore cannot help with choosing treatment. Further, many of the patients that do have actionable mutations will not respond to the treatment associated with that mutation. Only 5% of patients will benefit from genetic testing (4). Cancer cell mutations are only one part of a much larger picture, meaning they cannot capture the full complexity of why some patients will respond to treatment and others will not.
In addition to genetic testing, blood samples (also called liquid biopsies) have been used to measure a number of different components to better understand patient-specific cancer and predict the best treatment options for patients. These components include exosomes, peptides, proteins, and circulating tumor cells. Both the presence and quantity of each have been used to predict treatment response but few products are approved and on market.
Where precision medicine falls short
Both genetic testing and analysis of blood have improved the outcomes of patients. Through the identification of key mutations, new drugs have been developed and utilization of markers beyond genetics have allowed for some improved prediction of treatment response.
However, both tools rely heavily on historical data and make broad inferences based on an incomplete signal (cancer cells, blood components). These diagnostics predict how patients will respond but with a low degree of certainty while missing the bigger picture: the context in which the cancer cells exist. We need a better tool for determining the best treatment for every patient.
How we’re finding the best treatment for every patient
At Known Medicine, our custom micro-tumor platform brings together cutting edge biology research and the newest AI techniques to match every patient to the most effective treatment.
Cultured micro-tumor platform We use patient tumor samples to create hundreds of small tumors. We give these micro-tumors everything they need to behave the way they would in a patient’s body, including using immune and scaffolding (or stromal) cells from the cancer patient and extracellular matrix components found in the patient’s tissue. Then, we treat each sample with different drugs and drug combinations to see which works best for those tumor cells. From just one tumor sample we are able to generate over 300 micro-tumors and test 30+ treatments (chemo-, targeted-, and immuno-therapies).
Three-dimensional tumor models have been historically challenging to bring to patients because of the complex scientific methods required. To obtain patient cells from a tumor sample they must be removed from the natural tissue scaffolding and placed into a specialized culture plate or 3D scaffold. The new environment in which the cells are placed is key for preservation of the mixed cell populations and maintaining viability of the overall culture. The 3D scaffolds are complex formulations, often of natural and synthetic materials, which further complicates creating ideal models. However, advancements in biomaterials and cancer tissue engineering have recently allowed for such cultures to be viable in the lab. Our platform takes advantage of and combines these advancements with state-of-the-art high-throughput automation to help determine the best treatment for every cancer patient.
Custom analytics Once the micro-tumors have had some time to respond to the treatments, we can take a look at the results. Cancer cell death is just one factor here — we use machine learning to look for phenotypic indicators of real patient sensitivity, many of which are not clear by eye. These readouts are made available to oncologists in less than 7 days after tumor removal (several weeks sooner than genetic testing results) and can be used as decision support tools to determine the best treatment for every patient.
Our advantage We are driven to create a platform that best recreates the patient’s tumor while maintaining the ability to scale. We use novel biomaterial to make our platform automation friendly while advancing the capabilities of high-throughput modeling. We do all of this without sacrificing the robust tumor environment required to recapitulate the entire tumor, not just the cancer cells. The care we are taking to generate the best patient-specific data doesn’t stop at the 3D micro-tumor model — we ensure data is collected without bias to enable purpose-built AI models, ultimately allowing us to learn about cellular interactions and make clinically relevant insights. Our goal is to revolutionize every step of oncological care, from selecting the most effective treatment for each individual to developing the best new treatments for patients faster.
Siegel, RL, Miller, KD, Fuchs, H, Jemal, A. Cancer Statistics, 2021. CA Cancer J Clin. 2021: 71: 7‐ 33. https://doi.org/10.3322/caac.21654
Bailey MH, Tokheim C, Porta-Pardo E, Sengupta S, Bertrand D, Weerasinghe A, Colaprico A, Wendl MC, Kim J, Reardon B, Ng PK, Jeong KJ, Cao S, Wang Z, Gao J, Gao Q, Wang F, Liu EM, Mularoni L, Rubio-Perez C, Nagarajan N, Cortés-Ciriano I, Zhou DC, Liang WW, Hess JM, Yellapantula VD, Tamborero D, Gonzalez-Perez A, Suphavilai C, Ko JY, Khurana E, Park PJ, Van Allen EM, Liang H; MC3 Working Group; Cancer Genome Atlas Research Network, Lawrence MS, Godzik A, Lopez-Bigas N, Stuart J, Wheeler D, Getz G, Chen K, Lazar AJ, Mills GB, Karchin R, Ding L. Comprehensive Characterization of Cancer Driver Genes and Mutations. Cell. 2018 Apr 5;173(2):371–385.e18. doi: 10.1016/j.cell.2018.02.060. Erratum in: Cell. 2018 Aug 9;174(4):1034–1035. PMID: 29625053; PMCID: PMC6029450.
Marquart J, Chen EY, Prasad V. Estimation of the Percentage of US Patients With Cancer Who Benefit From Genome-Driven Oncology. JAMA Oncol. 2018;4(8):1093–1098. doi:10.1001/jamaoncol.2018.1660