Digital Solutions In Cancer Care

The development of effective methods for the diagnosis, treatment, and prevention of cancer is a significant challenge for the modern healthcare system. In the previous article, we analyzed the methods of AI-based imaging diagnostics of cancer. Let’s discuss the role of healthcare software solutions in the development of such highly demanded areas as liquid biopsy, predictive analytics, and personalized treatment and care.

Innovative cancer blood tests backed by tech giants

Blood tests for tumor markers known as prostate-specific antigens (PSAs) are a widespread method of cancer screening. In this sense, a liquid biopsy is a revolutionary AI-enhanced solution for identifying new tumor markers for screening. Blood tests look for cells shedding from a tumor into the bloodstream and/or circulating pieces of tumor cells’ DNA.

Grail and Freenome - startups from the USA with a total investment of $2 billion and $507.6 million correspondingly - have achieved success in this area. Both have relationships with tech giants - Freenome got investments from Google-related Alphabets Verily, and Grail is backed by Amazon.

The idea behind Grail is to build a highly accurate multi-cancer early detection blood test through smart processing of data on the presence of cell-free DNA (cfDNA) that is released by the tumor cells into the blood. Currently, the sensitivity of the new test in revealing 12 pre-specified cancers that account for approximately two-thirds of annual USA cancer deaths is an impressive 67.6%, and research is continuing.

Freenome focuses on solutions for early diagnosis of colorectal and prostate cancer through the integration of data on cfDNA with other circulating blood-biomarkers. In the pilot research of an ML-augmented screening blood test, screening from Freenome for early-stage colorectal cancer showed an impressive 85% of sensitivity and specificity. If the effectiveness of this approach is confirmed in further clinical trials, it could revolutionize colorectal cancer screening. Today, screening is carried out through periodic colonoscopies - an instrumental intervention that carries a disproportionately greater risk of complications.

Personalizing cancer care

According to the US National Cancer Policy Forum delivery of the right treatment to the right patient at the right time is a hallmark of high-quality cancer care. This approach requires the development of personalized treatment regimens and recommendations for patient care, taking into account possible risks and prognoses.

The use of IT solutions in personalizing cancer treatment can be exemplified by immunotherapy. Its effectiveness is determined by the patient's individual predisposition. The undoubted advantage of immunotherapy is its gentle nature and a small number of side effects, and the disadvantage is its high cost (about $200,000 per year). Combining AI algorithms with CT scan data helps to successfully identify groups of patients in whom immunotherapy will give the maximum effect. Personalized immuno-oncology allows for a sensible consumption of healthcare resources.

Technologies with AI personalize highly toxic radiation therapy. For example, a computer model makes it possible to better assess the likelihood of side effects, increasing patient safety. Siemens-funded AI developments by The Cleveland Clinic incorporate a patient’s EHR data array with a personal history of the disease, imaging studies, and laboratory research data into the analysis. This makes it possible to accurately assess the likelihood of a successful radiotherapy regimen for each patient.

Applying AI helps to better predict the risk of certain events in the course of the disease, such as readmission, recurrent oncology, treatment response rate, the potential toxicity of treatment, and the likelihood of mortality. This is a promising way to improve the methods of cancer therapy, even though these technologies have shortcomings as well. They raise questions about the representativeness of databases for training AI algorithms and the effectiveness of their treatment recommendations, even if they are developed by such large companies as IBM.

Personalization of cancer care is developing mainly through cooperation between clinical facilities and private or state innovation funds. The variety of cancer types and treatment methods opens significant opportunities for the cooperation of startups and software vendors with clinical and research institutions in creating AI-based clinical decision support systems. Such solutions take cancer care to a brand new level.

Enhancement of clinical trials

If all the treatment options available in clinical practice failed to help, doctors resort to experimental therapies that are in the clinical research phase. If the clinical situation develops in an unfavorable direction for the patient, digital health solutions come to the rescue:

• Mendel.ai from the USA (investments of $25.4 million) - the development of software solutions for oncology in the field of EHR data analysis and their coupling with existing research results and real-world evidence data.
• TrialJectory from the USA (investments of $7.7 million) offers an AI-powered decision assistance platform that ensures smoother access to clinical trial search and enrollment.

Patient support and monitoring

An example of an advanced system for remote cancer patient monitoring and supportive care is the NIH-funded CYCORE (Cyberinfrastructure for Comparative Effectiveness Research). This solution was developed at The University of Texas MD Anderson Cancer Center in collaboration with UC San Diego and The University of Alabama. The system is aimed at smooth data exchange between cancer patients and sensors at their homes, healthcare providers, and researchers.

Studies have shown that patient portals and apps play a significant role in promoting self-management and increase the longevity of cancer fighters. An example of an effective digital solution is a web-based symptom monitoring tool from Memorial Sloan Kettering Cancer Center (USA), which allows chemotherapy patients to survive longer.

Digital solutions for supporting cancer patients are also developed by other leading universities and clinics around the world. Distress Assessment & Response Tool or DART (Canada) makes patient-centered standardized validated measurements of physical distress, anxiety, depression, and social difficulty. Symptom Care at Home (USA) developed a solution for obtaining the patient's perspective on their health experience, using patient-reported outcomes. The British NHS-recognized startup Vine Health (investments of £1.2 million) is supplying healthcare providers with intelligent digital technologies for highly personalized patient support programs by using a combination of behavioral science and AI.

The symptom self-management platform Theraxium Oncology driven by evidence-based algorithms deserves a separate mention. This solution was built by the digital therapeutic company Voluntis in collaboration with leading pharmaceutical companies (AstraZeneca, Novartis, and Bristol-Myers Squibb).

The number and variety of companies and startups working in the field of monitoring possible symptoms of cancer and supporting patients is truly vast. Here are some of them:

• SkinIO from the USA - a developer of devices for skin health monitoring and early diagnosis of cancers via AI analysis of clinical-grade full-body photography of a patient. Despite the existence of numerous algorithm-based photo-processing apps for diagnosing melanoma and other types of skin cancer, we haven’t achieved an acceptable level of confidence in the detection of all cases of melanoma or other skin cancers when using these applications yet. Hence work in this direction should be continued.

• AI Talos from Turkey, which is developing AI algorithms for analyzing Infrared Thermal Imaging of mammary glands - a harmless tool for women’s health monitoring and breast cancer screening.

• Health Care Originals from the USA - the producer of a flexible patch-like wearable applied to a patient’s chest and tracking their coughs, respiratory pattern, wheezing, as well as detecting heart rate, skin temperature, and more. The device is connected with a patient app and portal and can be used in pulmonary patients, including patients with lung cancer. The solution has already proven itself worthy in the area of predicting obstructive chronic lung diseases.

Not all areas of cancer care are developing equally quickly, and not all technologies being developed have reached the stage of widespread adoption; however, it is digital technologies that help them progress faster.

Today, research, government, and business collaboration prevails in the development of some cancer care tools for genetic research, personalized care, and advanced cancer therapy based on Deep Learning. This entails a corresponding level of funding. On the contrary, the interests of private companies are mainly realized in creating solutions for Predictive Medicine, TeleMedicine, and complementary care. They aspire to promote products to promising markets and increase the variety of offerings and user comfort.

Cancer care remains one of the most sought-after and investment-attractive areas for the development of digital technologies in healthcare. There are still many areas in cancer care where the joint efforts of startups, MedTech and healthcare software development companies, scientific institutions, and innovative foundations can preserve the health and prolong the life of cancer patients.