Reboot Rx’s COVID-19 & Cancer Resource

Reboot Rx, formerly Cures Within Reach for Cancer, is a nonprofit developing new and affordable cancer treatments more quickly than traditional approaches. Our technology uses AI and machine learning to aggregate massive amounts of data on repurposed generic drugs and reduce the time to identify the most promising treatments from years to weeks. 

Our solution is a tool that leverages our evidence synthesis technology to inform patient decision-making. Patients with certain cancers are nearly three times as likely to die of COVID-19 compared to those without cancer. With the exponentially growing body of evidence on potential COVID-19 treatments (many are repurposed generics) and their effects on cancer, patients and doctors have to manually review thousands of studies on these drugs. Our COVID-19 & Cancer Resource is a web application that helps cancer patients who contract COVID-19 quickly assess how a particular treatment choice might affect both their COVID-19 and cancer.

Each year worldwide, 17 million people are diagnosed with cancer, $1 trillion is spent on their care, and 10 million people die from the disease. Cancer patients, especially those with COVID-19, need more effective and affordable treatments. Yet it takes 10 years and a billion dollars to develop one new drug, which then costs around $100,000 a year per person. Meanwhile, 106 repurposed generic drugs have shown promise for treating COVID-19 and are now being evaluated in clinical trials. Separately, 37 of these have also shown promise for treating cancer. 

Cancer patients with COVID-19 struggle to determine the best treatment options as they are either unaware of or overwhelmed by how to find and evaluate research on the effects of potential treatments. They rapidly need to make decisions about which treatments to take for simultaneously fighting both deadly diseases. With sufficient evidence, repurposed generic drugs could be used immediately to extend and improve their lives. New studies on these repurposed generic drugs are being published rapidly across multiple sources and it is impossible for patients and doctors to keep pace. They urgently need information that will help them evaluate the existing evidence in order to determine potential treatments.

Our solution is a web-based application that helps cancer patients who have contracted COVID-19 make informed treatment decisions by aggregating and presenting data on drugs. Our technology uses large manually curated biology training datasets and natural language processing pipelines to conduct intelligent evidence review at scale. We recently started applying this framework to synthesize cancer-relevant data for the drugs most actively being researched for COVID-19. Most of these are repurposed generic drugs, such as dexamethasone, a generic steroid that was recently proven to save the lives of patients seriously ill with coronavirus. We identified around 4,000 published clinical and preclinical studies that may contain evidence on the effects of the top COVID-19 drugs on cancer. We can now use our machine learning algorithms to find the subset of studies that are actually relevant and extract certain information from the relevant studies, such as the type of cancer that the drug was tested for and whether the drug was found to be beneficial or detrimental. We will then release this data and the insights gleaned in a web application that will quickly help patients and doctors make treatment decisions.

Our target population is cancer patients who have contracted COVID-19. Patients with certain cancers may be three times as likely to die of COVID-19 compared to those without cancer. Already fighting one deadly disease, COVID-19 is an additional hurdle that cancer patients and their caregivers have to face.

It would be impossible for doctors and patients to review the thousands of clinical and preclinical studies that describe the effects of candidate COVID-19 drugs on cancer. Our technology will do the heavy lifting and sifting so that patients can focus on the drugs that have the most promise to treat both cancer and COVID-19. 

We are interviewing cancer patients to develop our web application with functionality best suited to serve their needs. Cancer disease foundations represent an important constituency as well. Their work on behalf of patients afflicted with various types of cancer could be greatly supported by our technology platform. This summer, we engaged a team from MIT who will interview cancer disease foundations to better understand their approaches to researching treatment options and to learn how our technology could assist their efforts.

During infectious disease outbreaks, cancer patients are an especially vulnerable population. Cancer patients with COVID-19 need to rapidly decide which treatments to take that will give them the best chance of fighting both cancer and COVID-19. Our technology analyzes data on repurposed generic drug treatments that will inform patient decision-making. Our publicly available COVID-19 & Cancer Resource web application will share data on the most promising repurposed generic drugs with cancer patients around the world. This actionable information could immediately improve their lives, as repurposed generic drugs are affordable and have widespread availability.

There are a few organizations working on either building evidence synthesis technology or repurposing generic drugs. We are not aware of any that aim to combine these efforts to improve outcomes for cancer patients, especially cancer patients with COVID-19. For example, the Anticancer Fund is a nonprofit in Europe that offers evidence-based information about additional cancer treatments, including repurposed non-cancer generics, to patients who want to make informed decisions. However, their work relies on manual literature review, which is laborious and does not allow for identification of the most promising drugs in a timely manner. Efficiency in evidence review is of paramount importance during a pandemic. Our AI-powered engine works rapidly at scale to dynamically integrate new data into our platform and prioritize treatments in real-time. 

Few oncologists are even aware that many of the non-cancer generic drugs being evaluated for COVID-19 have shown evidence of having an effect on cancer as well. By using machine learning to focus on the effectiveness of generic drugs to treat cancer and COVID-19, we have the ability to quickly identify and disseminate the information that patients and doctors need to battle both diseases simultaneously.

Our technology uses AI to synthesize massive amounts of data on repurposed generic drugs to conduct intelligent evidence review at scale and then produce a ranked list of cancer-drug pairs. At the core of our product are machine learning algorithms to synthesize different data types: unstructured text data from published studies, structured clinical data from electronic medical records, and other patient and physician-reported data. Insights derived from these analyses will be communicated to a diverse set of beneficiaries through distinct web applications designed for their specific needs. We are setting up the backend infrastructure and data pipelines for our MVP to enable real-time inference of published studies using the best performing models. We have started designing UI mockups for a web application geared to share insights, particularly a prioritized drug list, with patients and doctors. We plan to develop and deploy a prototype of this web application after finalizing user requirements in the coming months.

Our initial models are already able to predict whether a scientific publication contains relevant evidence related to cancer treatment with 85% accuracy, and then predict whether the drug discussed was found to be effective with 73% accuracy. We published our results and presented at leading AI and machine learning conferences, including the Annual Conference on Innovative Applications of Artificial Intelligence in February 2020.

Our theory of change begins with the goal of improving the lives of cancer patients. We can do this through a series of steps targeting three separate problems: 

Cancer Patients need more treatment options faster

• Activity - Use AI and machine learning to rapidly identify the most promising non-cancer generic drugs for repurposing

• Output - Produce rankings of cancer-drug pairs

• Outcome - Sufficient data reporting the efficacy of non-cancer generics

Lack of knowledge that non-cancer generics could be helpful to cancer patients

• Activity - Create publicly available web application, Publish and report findings through media, conferences, and other outlets

• Output - Usage of website, Self-reported utility of repurposed generics by patients and doctors

• Outcome - Broad-based understanding of generics that can provide benefit to patients

Lack of financial incentives to repurpose generic drugs for other indications

• Activity - Develop innovative models to fund Phase 3 clinical trials where needed

• Output - 10 clinical trials conducted on promising generics

• Outcome - Sufficient evidence available from clinical trials

Desired state - Adoption of repurposed generic drugs into standard of care

In the long-term, we envision our technology will be used to identify the most promising repurposing opportunities for many other diseases. Imagine if today’s generics could be repurposed to improve outcomes for those with Multiple Sclerosis or ALS. Furthermore, since these generic drugs are low-cost and available worldwide, the treatments could have a major impact even in the developing world.

We aim to help the 17 million people diagnosed with cancer each year and the subset who may contract COVID-19.

1. Within 2 months: We will release a publicly accessible website with data on the drugs most actively being tested or used for COVID-19, and initial insights derived from our evidence synthesis platform.
2. Within 8 months: We will generate a prioritized list of generic drugs based on the evidence of their potential efficacy. To this end, we need to develop our evidence synthesis technology to the point where multiple types of data on repurposed generic drugs are included as input and a ranked list of cancer-drug pairs is the output. We are focusing on building the underlying training datasets, algorithms, and scoring methodology to accomplish this. 
3. Within 12 months: We will get the word out about our work to cancer patients and doctors so that they can use the surfaced information to inform their decision-making. We will partner with disease foundations, and begin tracking and reporting on key metrics to help us improve the application and understand its impact for patients.

For some cancer-drug pairs, we expect there to be enough evidence from published Phase II clinical trials plus our real-world data to immediately change the standard of care without the need for additional trials. For cases where promising cancer-drug pairs are lacking definitive evidence of their efficacy, we are developing a new funding strategy, social impact bonds, to finance their clinical trials.

In the near term, we need financial support to hire additional scientific and technical staff for our COVID-19 project. This will help us to rapidly build the targeted application for cancer patients with COVID-19 and ensure that our web application interface is easy for patients and doctors to use.

Over the next five years, we plan to fill the gap created by a market failure for repurposing off-patent and inexpensive drugs. Due to the lack of profitability, there is no commercial interest in repurposing generics. This means that there is no funding for the definitive clinical trials necessary to change the standard of care for cancer and COVID-19 treatments. As our technology begins to reveal the most promising generics that could treat cancer and COVID-19, we will face an increased need to bring together different stakeholders, including governments, impact investors, and value-based healthcare payers, in order to fund these clinical trials.

In order to address the market failure for funding clinical trials evaluating the efficacy of repurposed generic drugs, we are completing the framework for using social impact bonds (also known as pay-for-success programs), a new financing strategy. Social impact bonds will be used to engage stakeholders and fund clinical trials for promising repurposed generics that do not have conclusive evidence published on their efficacy. Implementing this strategy will enable us to change the standard of care for cancer and COVID-19 treatments as quickly as possible.

• We have 4 full-time staff. 
• 21 intern fellows are working with us who have backgrounds in biology, public health, computer science, and economics.
• Our working Board of Directors and slate of advisors are critical to our growth. They provide specific assistance in science, technology, fundraising, legal, and organizational areas.

Our team is uniquely positioned to succeed due to our cross-disciplinary expertise in areas such as cancer drug development, AI and machine learning, technology, precision medicine, real-world evidence, and outcomes-based contracts.

• Founder and CEO Laura Kleiman, PhD in Computational and Systems Biology from MIT, worked as Scientific Research Director in the Department of Data Sciences at the Dana-Farber Cancer Institute. 

• Co-founder and CSO Catherine Del Vecchio Fitz, PhD in Cancer Biology and MS in Medicine from Stanford, led development of precision medicine software platforms at Dana-Farber. 

• Co-founder and COO/CTO Pradeep Mangalath, Masters in Biomedical Informatics from Harvard Medical School and MBA from UNC-Chapel Hill, has experience in data science, engineering, product development, technology, and business. 

• Co-founder and Board Chair Andi Pollinger, MBA from Simmons Graduate School of Management, spent 30 years in financial services in marketing strategy and business management.

Cancer has touched all four of us, and our personal experiences drive us as we embark on the ambitious initiative to create a better future for cancer patients. Our founder’s vision, motivated and informed by her personal connection to the disease through the loss of her mother, combined with her extraordinary acumen as a cancer researcher and computational biologist, gives us confidence that we have a sound approach for bringing more treatments to patients more quickly.

We are fortunate to collaborate with leading academic researchers, corporations, regulatory agencies, and nonprofits to accelerate progress on our technology and impact the lives of patients. For example:

• We work closely with the IBM Research Science for Social Good Initiative, which partners with nonprofits to leverage their AI expertise and machine learning models to solve global challenges. We have been able to build on their previous work where they developed automated pathogen feature extraction from scientific literature. 

• By collaborating with Professor Byron Wallace at Northeastern University, we have been able to make faster progress. Professor Wallace has developed machine learning algorithms that are helping us quickly synthesize clinical trial information on repurposed drugs. 

• We are working with Professors Stan Finkelstein and Roy Welsch at MIT to generate real-world data to supplement published study results. Through the use of electronic health record (EHR) and claims data, we are analyzing the use of repurposed non-cancer generic drugs by cancer patients.

• We also work with MIT through the Entrepreneurship Lab and Volunteer Consulting Group to conduct market research.

• We have partnered with the COVID-19 and Cancer Consortium, which consists of more than 80 leading cancer centers in the US. The focus is on rapidly collecting and disseminating information related to cancer patients who contract COVID-19.

• We are collaborating with the nonprofit Mission: Cure to apply our technology to other diseases (their interest is in repurposing drugs for chronic pancreatitis) and develop innovative funding strategies for clinical trials to test repurposed generic drugs.

In the near-term, we anticipate that our work will continue to be supported by large foundations and high-capacity philanthropists. We are building relationships and submitting grant applications where there is good alignment between our mission and their interests. We have also been fortunate to participate in accelerator programs, which increase our visibility with social impact funders in particular. 

Additionally, we intend to make our platform available, such as through licensing or a fee-for-service model, to other organizations whose missions align with ours - to find and develop effective and affordable treatments for a myriad of other illnesses. There is an entire cancer care and research ecosystem that provides multiple routes to helping people with cancer while offering ways to generate sustainable funding for our nonprofit. Last year, we identified several potential customer segments and determined disease foundations would be the best place to start. A group from MIT is now conducting additional market research with cancer foundations specifically so we can develop our technology to meet their needs while assessing potential pricing models. 

We eventually hope to expand our customer base to include value-based payers and providers focused on improving patient outcomes while lowering healthcare costs, and molecular profiling and pharmaceutical companies that can benefit from integrating our insights into their decision workflow. All proceeds will flow back in to further support the development of our technology platform and generation of information for the benefit of patients.

We believe in our ability to connect cancer patients to effective and affordable treatments for their cancer and COVID-19. Through the myriad of partners and advisors available to Solve teams around the world, Solve will be able to connect us with resources that will enable us to take our solution to the next level.

We need partners who can help us build awareness of our web application and help to make it available to patients and their doctors. Cancer disease foundations like the Multiple Myeloma Research Foundation and the Breast Cancer Research Foundation are ideal candidates. The World Health Organization represents another potential partner whose reach and influence could help us access cancer patients and their caregivers worldwide.

Reboot Rx’s Founder and CEO has witnessed first-hand the pain and grief that a cancer diagnosis brings having lost her mother to the disease while working as Scientific Research Director at the Dana-Farber Cancer Institute. Dr. Kleiman spent years developing her vision for a way to get effective treatments to patients faster. She started Reboot Rx (formerly Cures Within Reach for Cancer) to pursue her dream of improving the lives of cancer patients.

Developing new therapies from scratch is much less efficient than developing new uses for FDA-approved drugs that have already been proven safe. Repurposing generic drugs for new indications is a game-changing strategy to expand treatment options for patients. We aim to change the way patients and doctors think about treating cancer and infectious diseases by incorporating effective, safe, and affordable repurposed generics into the standard of care.

The Elevate Prize would supply us the tools we need to further our mission of increasing the accessibility and affordability of medicines. The Elevate Prize funds will help us accelerate the development of our project and allow us to reach patients in all corners of the world. Even more significant than the financial award will be the ability to raise the profile of our work with influential partners that can advance our project. The next two years are critical for Reboot Rx, and the support of experts in fields such as business, marketing, and technology will help us develop and scale our organization and its impact.

Reboot Rx is harnessing the power of AI and machine learning to improve the lives of cancer patients. Cancer is a horrible disease that not only affects those with the diagnosis, but also their families. The emotional and financial toll on patients and their families is devastating. AI is what makes our solution possible. It is the underpinning of our technology and makes it possible to reduce the time it takes to review and synthesize data from thousands of published research studies from years to weeks. We have started with cancer. But there are hundreds of diseases for which repurposed generic drugs could be used to improve patients’ lives all over the world. AI will allow us to scale our technology beyond cancer and continue to enhance the quality and precision of the insights our platform will provide.

Repurposing generic drugs to treat cancer and COVID-19 is an effective and affordable alternative to developing new drugs. However, traditional funding mechanisms are out of the question as there is not enough commercial interest in repurposing generics. Crowdfunding could greatly impact the progress of our project and ensure that we can deliver useful information back to the public as soon as possible.

Pharmaceutical companies are not investing in repurposing generics because they do not foresee making a profit off of these inexpensive treatments. Time is of the essence for cancer patients who contract COVID-19, and repurposed generics have the potential to improve the lives of people around the world much faster than new drugs. Taking the problem into our own hands, this more affordable alternative could be crowdfunded by individuals. The funds collected would be used to quickly complete our project in order for us to maximize our impact on the lives of cancer patients.

Eventually, we hope to expand the use of our technology and work towards repurposing generic drugs for diseases beyond COVID-19 and cancer. The People’s Prize will also assist us in advancing towards a future where all medicines are accessible and affordable to everyone.