Each patient’s cancer has a unique genetic make-up. Precision medicine aims to characterize these differences to target the right treatments to the right patients at the right time. Despite some major successes with targeted cancer therapies and immunotherapies, so far only a small percentage of patients have benefited from such treatments.
The US-based company SimBioSys aims to deliver precision medicine by developing software that creates an in silico model of a patient’s tumor and predicts its response to specific treatments. Founded in 2019, SimBioSys’ 4D tumor modeling platform, TumorScope, provides predictions that can inform individual treatment plans. It can also generate new insights into cancer to inform drug discovery efforts and optimize clinical trials using previously acquired standard-of-care data alone.
The platform integrates available patient data (including electronic medical records, genomics, imaging and pathology results) with proprietary AI and multiscale tumor simulations to accurately predict the behavior of a tumor and its response to therapy.
“By incorporating major hallmarks of cancer into a comprehensive model, such as tumor microenvironment, perfusion, metabolism and physical traits, TumorScope captures the manifestation of tumor heterogeneity and biology,” explained Tushar Pandey, SimBioSys CEO and co-founder. “Rather than going after the expression of just a few genes to understand tumor heterogeneity, our multiscale 4D approach allows us to predict more accurately how a tumor will evolve and respond to therapy over time.”
TumorScope helps clinicians and drug developers
The clinical effectiveness of TumorScope has been validated in more than 1,000 breast cancer patients across 25 clinical sites. Clinical studies have shown that TumorScope can predict patients’ pathological complete response (pCR) to neoadjuvant therapy in early-stage breast cancer patients with over 90% accuracy1. TumorScope can help oncologists make risk–benefit assessments of multiple treatment options for a patient’s tumor and can support their decision of escalating or de-escalating therapy based on predictions at baseline.
Importantly, the platform can support the drug-development life cycle. Researchers can use the platform’s predictions to assess why a patient is, or is not, responding to a therapy or combination of therapies. The in silico tumor model makes it possible to model metabolites, drugs and cancer cell response at voxel level, which allows researchers to model drug delivery and optimize dosing as they see how tumor biology changes over time.
TumorScope simulations generate a novel data set called simulomics, defined as the quantification of spatially resolved biological, structural and functional characteristics of a tumor. Simulomics can be used to identify responder populations and novel multimodal biomarkers of response that aid patient stratification and selection decisions in clinical trials. This information is crucial for the development of targeted drugs and immunotherapies that may benefit only a subset of patients.
The goal of SimBioSys is to aid decision-making with currently approved drugs as well as early-stage drug development, efficacy testing and the design of clinical trials. Newly approved drugs will become incorporated into treatment planning models to improve the patient journey (Fig. 1).
Fig. 1 | Using TumorScope to provide precision medicine improves functionality across the drug development life cycle. SVTB, SimBioSys’ Virtual TumorBank.
Using TumorScope for in silico clinical trials
SimBioSys has shown how TumorScope helps to bridge the gap between preclinical and clinical testing by computationally simulating clinical trials. Using SimBioSys’ Virtual TumorBank of more than 6,000 past solid tumor patients and accompanying clinical data, TumorScope has replicated the results of multiple clinical trials. These include the NeoSphere trial, which investigated the efficacy of combinations of docetaxel, pertuzumab and trastuzumab in HER2-positive breast cancer patients. The regimen efficacy, pCR rate and disease progression predicted by TumorScope closely mirrored those of the actual trial, highlighting how in silico trials can not only reduce the cost of clinical trials, but also describe potential detrimental effects on patient outcomes2,3.
Retrospective analysis of clinical-trial data using TumorScope has identified clinical features that distinguish progressors from responders. These features are being assessed to develop individualized predictive biomarkers of disease progression. “It is evident that single markers are not sufficient,” said Tricia Carrigan, SVP of Precision Medicine at SimBioSys. “To truly advance precision medicine, composite multimodal signatures capturing not just the molecular make-up but also the physical and metabolic characteristics of patients will be critical.”
SimBioSys is partnering with leading US cancer centers and pharmaceutical companies to continue the validation process and expand the use of its technology to all solid tumors. TumorScope is commercially available for research purposes, and SimBioSys is working towards achieving regulatory approval from the US Food and Drug Administration as the first medical device of its kind to aid treatment planning.