GNS Healthcare Launches Computer Model of Disease Progression, Treatment Responses in Myeloma

A computer program developed by GNS Healthcare could be used to model and predict disease progression and response to treatment in people with multiple myeloma.

Called “Gemini, the in silico ” Patient” (“in silico” refers to analyses done using computer simulation), the program was built through a collaboration between multiple biopharmaceutical companies, academic medical centers, and the Multiple Myeloma Research Foundation (MMRF).

Gemini builds on data from the CoMMpass Study, a project from the MMRF that is following more than 1,000 multiple myeloma patients several years after their diagnoses. Gemini includes data on many common myeloma treatment modalities (e.g., corticosteroids, proteasome inhibitors, CD38 inhibitors) and how they connect to various clinical endpoints (e.g., survival, disease progression). These data are then analyzed with artificial intelligence (AI) programs such that Gemini can “learn” — that is, identify connections and associations within complex data.

The overarching idea is that, having learned from the existing data, Gemini can predict the most likely outcome when given a new set of variables (patient characteristics, treatment(s), etc.) from a new patient.

“We are reaching a tipping point where patient data is becoming rich and multi-layered enough to power AI models that can help predict patient response at the individual level,” Ravi Parikh, MD, an oncologist and instructor at the University of Pennsylvania, said in a GNS press release. “This announcement represents a true step forward in personalizing cancer treatment.”

GNS highlighted multiple potential applications of Gemini: for instance, it could be used to run in silico clinical trials. This could be useful for designing actual clinical trials, as it could help identify individuals who are most or least likely to respond to a given treatment, allowing trials to recruit only participants expected to benefit.

Similarly, Gemini could aid in treatment decisions at the individual level. It can be used to predict the outcome of particular treatments, allowing for rapid trial-and-error that doesn’t involve the patient. Gemini also could help clinicians make decisions about the sequence in which treatments are given.

“Over the past decade there have been a dozen treatments approved for multiple myeloma but there is still a lack of evidence to ensure patients receive optimal treatments in first line and subsequent lines of therapy,” said Colin Hill, chairman and CEO of GNS.

“Creating Gemini, the in silico patient, allows us to break the bottleneck of understanding what treatments work for which patients, driving better clinical trial design, generating real-world evidence for market positioning and ultimately creating better outcomes for patients,” Hill said.