Driving Progress in Personalized Cancer Therapy
Treatment resistance to cancer therapy can be a major stumbling block in a patient’s clinical journey, but it can be addressed with combination therapy or personalized medicine.
Personalized medicine uses patient-specific information to tailor treatment strategies to each individual, reducing the likelihood of treatment resistance occurring. OncoHost uses proteomic analyses and AI-driven technology to give insights into how a patient will respond to treatment, equipping the patient with the information to put them in the driver’s seat of their care journey.
Technology Networks spoke with the CEO of OncoHost, Dr. Ofer Sharon, to find out more about how OncoHost enables patient-specific treatment plans based on personalized medicine approaches.
Katie Brighton (KB): Can you tell us about how personalized oncology treatments have developed over the last decade? How has this changed the patient experience?
Ofer Sharon (OS): Over the years, drug-based anti-cancer treatment has evolved from a non-specific ‘carpet bombing’ approach to targeted therapies – biologic therapies that target specific cancer cells that harbor ‘driver mutations.’ In addition, over the last decade there has been a rise in immunotherapies used in the clinic.
Immunotherapy is a new type of anti-cancer treatment that activates the patient’s immune response towards the cancer. Targeted therapies offer improved response rates with a different profile of adverse events, in that patients respond to treatment for longer, and are more likely to respond well. While this approach to cancer management improves the outcome of survival, immunotherapy also greatly improves clinical outcomes.
One issue with targeted therapies is that they are only relevant for a minority of patients for whom driver mutations can be identified. At some point during treatment, the patients may experience cancer resistance breakthrough and they eventually stop responding to their given treatment. With immunotherapy, a certain percentage of patients respond until they eventually experience this resistance breakthrough. This issue stems from the inability to pre-emptively determine which patient will respond and benefit from the course of treatment and which will not. This inability to predict patient response results in wasted time, unnecessary adverse events, frustration, despair and uncertainty.
KB: How does treatment resistance occur in cancer patients? Are there particular cancer types that are more likely to become treatment-resistant? How can we combat this?
OS: Resistance to treatment is a multifactorial process. Resistance occurs when cancer cells have a certain molecular trait that causes them to be resistant to a specific drug because of biological mechanisms related to the body’s response to the anti-cancer treatment. This phenomenon, known as host response, is one of the major reasons for treatment resistance today.
Resistance mechanisms are generally diverse and non-specific to certain cancer types and may present themselves as primary resistance, intrinsic resistance or acquired resistance. Combating resistance requires several strategies. Combination therapies, along with therapies that have different mechanisms of action, are well-known approaches used to mitigate treatment resistance. Another approach is personalized medicine. Personalized medicine is a specific approach that involves adapting the treatment plan to fit the patient’s specific needs rather than adopting a ‘one-size-fits-all’ treatment protocol.
KB: What benefits do proteomics analyses and artificial intelligence (AI) bring to understanding which treatment is best for the patient?
OS: Proteins are the building blocks and drivers of biological processes in our body. Analyzing proteins allows us to gain insight into the complex interplay of the tumor, the therapy and the host (patient). This complex biological interaction involves intrinsic cancer cell characteristics with the body’s host response to treatment and is the underlying reason for treatment resistance.
There are thousands of proteins in the body and making clinical sense of these protein levels and dynamics requires sophisticated mathematical and bioinformatic tools. Combining plasma protein analysis with machine learning tools enables us to answer three clinical questions:
- Will the patient respond?
- Why does resistance occur?
- What may be the next line of therapy?
The key factor of this combination analysis is that it is all completely personalized, specific and relevant to each individual patient.
KB: How does OncoHost help patients make decisions on their treatment plan?
OS: We provide oncologists with insight on the response probability for individual patients for the first year of treatment, analysis of the biological pathways involved in resistance and identification of potential resistance-associated proteins. We also provide analysis on the drugs that are targeting those resistance-associated proteins. This analysis allows us to intervene early and support the clinical decision-making process with a significantly improved level of insight.
KB: As personalized treatments are coming closer to being widely used in the clinic, are there any considerations that need to be taken into account regarding handling patient data?
OS: As with any other type of personal health information, we need to operate within the boundaries set by legislation and patient privacy standards and best practices. HIPAA and GDPR are good examples of those measures.
KB: Where can you see personalized therapy heading in the future?
OS: In the future, diagnostics is going to be based on several assessments across the continuum of the disease. Early detection, treatment guidance, identification of intrinsic resistance, acquired resistance and different treatment strategies are all examples of the future course of diagnostics. I believe that the data to identify these factors and guide clinical decisions will be based on ‘multiomic’ analysis at different points in time. This will allow for the guidance of clinical decisions that are specifically matched for each individual patient.