White Paper
Cancer therapy is undergoing a golden era of innovation.
Immuno-oncology (IO), which takes advantage of the body’s immune system for treatment development, is experiencing especially explosive growth. The FDA is currently overseeing 800 active applications for cell and gene therapies, and by 2025, could approve 10-20 applications each year.1 This continued trajectory is likely to categorically disrupt the entire healthcare system as we know it. IO innovation includes a range of approaches2 like immune checkpoint inhibitors also known as PD-1 or PD-L1, monoclonal antibodies, therapeutic vaccines, and adoptive T-cell (immune system cell) transfer, including chimeric antigen receptor T-cell (CAR-T) therapy, which can now be used to treat certain blood cancers.
The timing for this innovation could not be better: one-fifth of global pharma sales is spent on oncology,3 and 40% of U.S. adults will receive a cancer diagnosis in their lifetime.4 From clinical trial design, to manufacturing methods, to reimbursement and administration, innovative approaches are key to preparing for this moment. The flurry of biopharma gene therapy M&A activity in recent times indicates agreement from the biopharmaceutical industry. Gilead, for example, has made big bets on IO to build their cancer division. With over half a dozen deals totaling $27 billion, they have expanded their portfolio to include both experimental and marketed cancer drugs, signaling a clear focus on IO5.
This trajectory of innovation presents challenges, however.
For example, FDA-approved CAR T-cell therapies are expected to increase from two in 2019 to 24 in 2024.6 And that is only one IO treatment category. With the options changing frequently, how do oncologists match patients to the right treatment?
There are new data-related challenges as well. Far more data are becoming available to oncologists, caregivers, insurers, biopharma companies, and other stakeholders, making it more difficult to parse the data that are actionable and valuable. A wide variety of IO combination therapies are gaining approval for the treatment of a wide variety of conditions.
As the data sets for each combination become smaller, especially in late-stage clinical trials, validating results with sufficiently powered studies becomes more difficult. Novel therapies which may be more complex and carry new risks often lack remote monitoring services, limiting access to major cities and Academic Medical Centers.
In addition, high price tags for these IO therapies, often in the range of hundreds of thousands of dollars, requires manufacturers to show unequivocally improved outcomes in real-world data to continue and expand reimbursement from payers.
In this white paper, we walk through the role that digital is playing in helping to realize and accelerate the impact of IO therapy. In particular, we focus on:
The more data collected, the more information that can be acted upon. A regulated digital health platform enables transferring, processing, and analyzing data to support therapy development, making the data more accessible. To maximize IO therapy, it is critical to leverage data sets to build algorithms that will provide greater impact on cancer patients – from initial diagnosis all the way through to treatment monitoring.
Kaiser Associates estimates that by 2025, 70% of clinical trials will incorporate sensors, up from 10% to 15% in 2019.7 Connected wearables could get therapy management out of the clinic and into the outpatient setting faster, so patients can use remote monitoring and go home earlier.
Wearables can enable active communication between the patient and the care team, helping the patient feel connected. That connection can make for a better patient experience, which in turn can improve outcomes." — Oncology executive at a leading life sciences company
There are two categories of wearables: true wearables that track heart rate and other biometrics, and tools to track symptoms. Certain combinations of symptoms can signal disease progression and trigger earlier intervention, through prediction with AI algorithms.
Remote patient monitoring utilizing an app allows the physician or treatment team to periodically check in with their patient, using text for instance. This gives the patient a feeling of connection and may reduce the number of required in-clinic visits. In addition, this information from interactions and monitoring of the patient, gathered passively or actively, can help anticipate potential adverse reactions.
Remote patient monitoring is also enabling a shift away from Academic Medical Centers (AMCs) to community care centers. Patients would no longer have to travel far from their homes and support networks for extended periods of time in order to receive treatment. Being close to AMCs has traditionally offered the benefit of getting the best medical advice. However, technology can offer a democratizing effect, where patients in suburban or rural areas can still have access to similar medical expertise. The COVID-19 pandemic has made it more challenging and riskier for cancer patients to visit hospitals, further reinforcing the benefit of remote monitoring tools that can gather information and maintain a connection between patient and care teams. More can now be accomplished at local hospitals and at home, improving patient outcomes in a more comfortable and cost-effective manner.
Oncology is experiencing a significant increase in the volume and types of data coming from disparate sources including Electronic Health Records (EHRs), wearables, connected devices, and patient-entered information. Better clinical decision support tools are needed to help physicians sort through data and spot relevant trends at the point of care. Easy, timely access to that information, including integration with the EHR, is necessary to make informed care decisions, along with empowering the care team with the best treatment options.
The oncology patient journey is long and arduous, leaving numerous opportunities for digital intervention. However, it is clear that early interventions are most impactful. Physicians can utilize biomarkers and imaging data to stratify patients and predict which patient subtypes are most susceptible to experience poorer outcomes. Lack of these indicators early on can lead to breakdowns in patient care.
Data is pretty good in the U.S. but it depends on the ability of the organization to bring all of those data sets together so that they can actually speak to each other. Then you're able to interrogate that data in a manner which gives you insights. If you are able to do that, you have a quantum leap. Use these insights to provide and facilitate good conversations on things that are relevant.” — Oncology executive at a leading life sciences company
In the future, machine learning is trending towards taking existing data sets and running the tests again without having to wait for a new clinical trial. A Class III Software as a Medical Device (SaMD) for patient-specific treatment selection could be created just from existing data sets. This is especially helpful as therapy selection decisions get more complex with more candidate treatments and combinations. How hard is it going to be to bring these combinations to market and how can we streamline combination selection?
Many health IT systems, wearables, and biopharma companies run on proprietary software that does not integrate well with others. Given the data silos that exist between biopharma and diagnostic companies, oncologists are often missing the chemistry data and other clinical data that would provide a more holistic patient view. This greatly limits the value derived from data. Having a common underlying infrastructure that enables interoperability and integration across systems is critical.
Further, several stakeholders need matching data to treat each patient – radiologists, clinicians delivering chemotherapy, oncologists, primary care physicians, social workers and others, who are in different locations. Each has different requirements and uses for overlapping sets of data. These need to be coordinated for better patient outcomes.
As we get access to more data, one question to ask is, ‘can we build networks and understandings aggregating this information?’ We can look to connect the dots across different sites to follow one patient and build interaction maps.” – John Lo, Senior Vice President, Worldwide Hematology, BMS
Multiple data sets drive the need to build an algorithm to connect the dots and predict who will have the best results with a particular therapy. For example, there are fast progressors that chemotherapy can control, but just for six months. Questions remain on whether patients on a PD1 or PD-L1 therapy enhancer will continue to respond well. As of now, there are not good data sets to build an algorithm because factors are widely varied.
Pharma and the diagnostic companies have completely different need sets and development timelines. Also, the commercialization of a diagnostic assay is very different from the commercialization of a drug in the immuno-oncology space.” – Joe Bernardo, Operating Partner, Linden Capital. “We need to think about the average oncologist and get them activated.”
By better predicting how a patient’s cancer will react to a particular treatment, identifying what adverse events may occur and what combination of therapies is best suited for the patient, you can improve outcomes. Data from companion digital diagnostics, remote monitoring devices, and algorithms can help physicians make more informed treatment decisions.
Increasing the accuracy of patient matching – by getting the right patient the right therapy and maximizing adherence —could have greater impact on clinical outcomes than more primary innovation with new therapies.
A wide swing in patient outcomes calls for more accurate patient identification and matching to studies. The Kaplan-Meier curve is a popular method in clinical research for estimating probability of survival. However, censored data, including when a patient drops from a study or survives beyond the research period, significantly impacts findings.8 Digital offers better identification of those patients who will respond to a specific treatment and enhanced data integration for patients using multiple treatments.
Presently, mapping patients to the most appropriate FDA-approved treatments is challenging.
For instance, measuring PDL or PD-L1 on patient tumors only provides data at a specific point in time. If the patient is receiving chemotherapy, the PDL readout can change. Complementary diagnostics or guidelines can enhance this process.
If we could crack that, response rates would increase because you'd be selecting those patients, based on the features of their tumor and their immune response, who are more likely to have a good outcome. Using digital technologies, you could even help interrogate that data over time, incorporating different markers, or use the data in a different way that would be hugely helpful." – Teri Foy, Senior Vice President of Immuno-Oncology and Cellular Therapy at BMS
Better identification of and information around predictors of resistance, including tumor biopsies and blood work, can help connect patients with better treatments. Tissue samples before and after therapy would help but are difficult to obtain. As blood tests become more sensitive, this can help address these limitations. More informed development of IO therapies and better patient selections for treatment will follow. These are the early days for personalized algorithms.
One important area is identifying what is driving resistance to a particular treatment to identify the next therapy for that type of post-IO exposure.” – John Lo, Senior Vice President, Worldwide Hematology, BMS.
Informatics and predictive sciences can help bring all that information together, whether it is around cell therapies, drug product characteristics or translational data being sampled during the trials. Digital tools can improve matching patients to therapies and predicting adverse reaction, with the growing use of AI- and machine learning-based algorithms predicting who will respond to treatment or not.
Bringing clinical patient features together with response and biomarker data helps predict correlations around which patients may respond better and may allow us to identify features that we want to engineer into our therapies.” – Teri Foy, Senior Vice President of Immuno-Oncology and Cellular Therapy at BMS
While not validated as complementary or companion diagnostics, wearables do provide valuable insights into patient responses. Obtaining more biometric markers in the composite score would offer valuable support. Taking data points and pulling them together can define either correlations in responses or predictors of other patients that will respond.
We need to think about the patient journey. We should work toward treatment tools that capture the patient journey more holistically and go beyond a specific treatment modality. This may help to reduce the current fragmentation and improve the patient journey.” — Oncology executive at a leading life sciences company
It is also critical to ensure that treatment is introduced at the right point in time.
Making the wrong initial selection early on can have a massive impact on the patient’s life expectancy, so you want to get it right from the get-go. An initial decision at the start can have huge ramifications.” – Oncology executive at a leading life sciences company
Oncology is at the forefront of precision medicine, with more than 160 oncology biomarkers approved as of 2019 and more than 90 percent of pivotal trials against specific molecular targets.9 More combinations are being studied in clinical trials and will be approved. There will be panels of tests and there are more variables that get into the mix.
A key tenet for us is to follow the science. We know that tumor biology is complex and that patient response to drugs might differ based on one or more biomarker levels. So it is important to tailor treatment in order to get the best outcomes for patients. We are focused on advancing precision medicine.” – Oncology executive at a leading life sciences company
Significant development is driving more individualized treatments. Oncology is getting better at characterizing the tumor. That understanding can be applied at a nuanced level to an algorithm to determine what treatment would benefit the patient. This also complicates clinical trials.
A digital approach can be taken to develop an algorithm that would recommend which molecular tests should be run, upload the results and then offer the oncologist a suggested incremental test to determine the best therapy. That information can be coalesced to determine the best therapy, which may not be chemotherapy. Digital and AI-based algorithms can play a vital role in real-time, supporting oncologists with choosing specific testing protocols to determine the right treatment path. This is particularly helpful with fast-moving developments changing the IO landscape.
Streamlining the education and then the execution by every oncologist is where the value will be created with digital." – Joe Bernardo, Operating Partner, Linden Capital
After a patient has been matched to a treatment, the next hurdle involves reducing risk and optimizing outcomes. Remote patient monitoring and companion digital solutions can assist in determining patient reactions prior to treatment.
The most common side effects of CAR-Ts are cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS). Physicians are typically only able to intervene when the adverse event has progressed – a key driver for lengthy in-patient stays. Outpatient use of CAR-Ts comes with burdensome infrastructure requirements, including needing to closely monitor patients and admit them quickly if their CRS or neurotoxicity begins to worsen.10
An early indication of CRS is fever – digital monitoring of a patient’s biometrics, including body temperature, may help physicians predict CRS sooner and take mitigating action swiftly with the correct therapeutic intervention. Predicting neurotoxicity is difficult and it typically occurs much later than CRS. While there are predictive factors of neurotoxicity, physicians have yet to have an accurate model or algorithm to predict the timepoint and severity of neurotoxicity – though there is plenty of ongoing research to ascertain which patients may not benefit from an IO therapy. Researchers at MD Anderson found that monitoring early changes in circulating tumor DNA may predict patient response to CAR-T cell therapy.11 Remote monitoring could enable cognitive testing every few hours to identify early stages of neurotoxicity or other adverse events.
On the research side, digital can enable regular data collection from the patient throughout the patient journey. This would provide information on all side effects or adverse reactions regardless of severity and before they become severe, enabling physicians to make timely treatment decisions.
Data can be taken from blood samples to identify biomarkers earlier and provide deeper data to determine floating tumors or predict adverse events. Patients could receive treatment at their care setting and then go home with continuous monitoring data fed back to the care team, improving patient experience while reducing costs.
Moving forward, digital has the potential to accelerate trials, improve study participant recruitment and retention with earlier identification of cancer and better patient matching. Remote patient monitoring enables more home-based treatment and participation. These applications of digital health have the potential to significantly reduce the costs associated with clinical trials, increase patient enrollment and expand clinical trial coverage.
Meeting study participants where they are will make it easier for patients to enroll in clinical trials. One of the considerations that I’m becoming more mindful of is we tend to have study participants who are healthy and financially resourced enough to be in a trial. There’s a lot to improve upon so that we can enroll more heterogeneous populations: those without access to transportation, those across a wider range of socioeconomic status and literacy. Digital can help across all those areas." – Mintu Turakhia M.D. M.A.S., Associate Professor and Executive Director, Center for Digital Health, Stanford University and member, BrightInsight Clinical Board
Clinical trial enrollment is an expensive and time-consuming process. A cancer patient will learn about a trial from the treating oncologist. Typically, the oncologist suggests that a patient participate in a trial at their institution. Only 8.1 percent of oncology patients participate in clinical trials12, due in part, to the burdensome task of identifying a clinical trial at another institution or geographic area.
Artificial intelligence (AI) is assisting in the patient matching process by structuring patient records for ClinicalTrials.gov and comparing clinical trials inclusion and exclusion criteria. In-depth protocols reduce rejection rates by applying machine learning to pre-screen information against protocols.
COVID-19 has exacerbated existing challenges with clinical trials. Currently, requirements of a clinical trial participant are often more stringent than that of a patient receiving standard treatments, possibly requiring more in-person hospital visits for safety monitoring, imaging, and additional laboratory testing. These requirements present concerns for patients, institutions, and sponsors during a pandemic.13 The number of oncology clinical trials open for recruitment significantly decreased during the pandemic. Between February and May 2020, 920 clinical oncology trials were interrupted due to COVID-19.14 A search in ClinicalTrials.gov using the keyword “cancer” and filtering for clinical trials with a recruiting status showed a 41.8% reduction when comparing two time periods: January to May 2019, and January to May 2020. From January 1, 2020, to May 31, 2020, 817 clinical trials were found, compared with 1405 during the same period last year15. Increased use of remote monitoring of patients could mitigate the impact to clinical trial enrollment and progress during future public health crises.
COVID-19 has exacerbated existing challenges with clinical trials. Currently, requirements of a clinical trial participant are often more stringent than that of a patient receiving standard treatments, possibly requiring more in-person hospital visits for safety monitoring, imaging, and additional laboratory testing. These requirements present concerns for patients, institutions, and sponsors during a pandemic.13 The number of oncology clinical trials open for recruitment significantly decreased during the pandemic. Between February and May 2020, 920 clinical oncology trials were interrupted due to COVID-19.14 A search in ClinicalTrials.gov using the keyword “cancer” and filtering for clinical trials with a recruiting status showed a 41.8% reduction when comparing two time periods: January to May 2019, and January to May 2020. From January 1, 2020, to May 31, 2020, 817 clinical trials were found, compared with 1405 during the same period last year15. Increased use of remote monitoring of patients could mitigate the impact to clinical trial enrollment and progress during future public health crises.
How do we give time and freedom back to patients instead of them planning their life around treatment schedules? Digital can give more flexibility, freedom and a more sophisticated regimen.” – Oncology executive at a leading life sciences company
Outcomes data will start opening reimbursement for digital, slowly. Digital health will also better position healthcare organizations to take advantage of another reimbursement trend that favors outpatient over inpatient therapies.16 COVID-19 has helped this process as many payers began reimbursing for some remote and virtual care, which may lead to more coverage as the benefits are documented. CMS has been placing a greater importance on home-based and value-based care – a promising sign of future reimbursement for tools enabling these pathways.
CMS is interested in value-based care and immunotherapies are extremely expensive. If digital can help to solve the question of who gets the most benefit out of this therapy, then payors may be more likely to reimburse for these treatments.” – Arturo Loaiza-Bonilla, M.D., MSEd, FACP, Cancer Treatment Centers of America
Appropriately leveraging digital health in immuno-oncology solutions will improve outcomes, enable more personalized interventions and increase access to benefit more patients. Through avenues including AI, wearables and remote monitoring, digital health can play a significant role in expanding data points and sources, improving clinical trials, preventing adverse events, enabling home care and improved quality of life.
In the future, what can digital do to ensure that IO delivers maximum survival and quality of life? A big differential exists between companies that understand this and those that are still figuring it out.
IO will continue generating meaningful impact for patients. Which biopharma and medtech companies lead the industry will largely depend on innovation and the rate at which they embrace digital to bring therapeutics to market faster and create competitive differentiation.
BrightInsight, the provider of the leading global regulated digital health platform for biopharma and medtech, is actively working in the immuno-oncology space. The BrightInsight™ Platform uses software and services to capture, transmit, and analyze data from CE-marked and FDA-regulated medical devices, combination products, apps and Software as a Medical Device, in compliance with global security, privacy, and regulatory requirements.
With the world’s leading biopharma companies as customers, BrightInsight’s compliant infrastructure is powering remote patient monitoring, advanced algorithms and digital clinical trials, and more immuno-oncology use cases.
Deployed as a managed service, the platform accelerates time to market for biopharma and medtech companies, reduces the cost of implementation and maintenance versus a custom solution, and scales across products and global markets.
Our team brings deep expertise across digital health leadership in the biopharma and medtech industries, including building and launching some of the very first cloud-connected and FDA-approved combination products, and hosting innovative SaMD and algorithms.
Our vision is to transform patient outcomes globally by bringing the power of digital technology to healthcare, and we work every day to achieve this by accelerating regulated digital health innovation for our customers through our scalable medical-grade platform.