Immunotherapy Weds Artificial Intelligence: All Cancer Patients are Invited
Healthful Vitality | 04/16/2019 | By Dr. B. R. Achyut, PhD | Edited 05/21/2021 | Immunotherapy Weds Artificial Intelligence: All Cancer Patients are Invited
Cancer is one of the most complex and devastating diseases in humans. According to the American Cancer Society (ACS) facts and figures of 2021, an estimated 1.9 million new cancer cases will be diagnosed in the United States. Among these, almost 608,570 patients will die despite available surgery, radiation, and chemotherapy options, which are considered the first line of standard care in cancer. In the past, several adjuvant treatments have been tested to improve patient outcomes in clinical trials. Still, most trials either failed or provided minimal to no survival benefits in diverse cancer types. Before describing immunotherapy weds artificial intelligence, let us examine immunotherapy and what artificial intelligence is.
Immunotherapy
A recent development in cancer therapy is immunotherapies. Immunotherapies improve the immune system to fight against cancer. Notably, this development led to a Noble Prize in Medicine in 2019. Immunotherapies such as targeting immune checkpoint inhibitors (PD-L1 and CTLA-4) hold promise for cancer therapeutics. In addition, dendritic cell therapy, cancer vaccines, and CAR-T cell therapy have also boosted great enthusiasm among physicians and cancer patients after remarkable benefits.
However, immunotherapy cannot be generalized in all patient cohorts because it is only successful in a subgroup of patients. Also, reports show the development of immunotherapeutic resistance in patients. This resistance indicates the need for combination therapy. Nevertheless, immunotherapy is a good option. And it is FDA approved for melanoma, triple-negative breast cancer, lung cancer, kidney cancer, Hodgkin lymphoma, and urothelial carcinoma, a common type of bladder cancer. In addition, immunotherapies are emerging as an option for several late-stage cancers. In the future, chemotherapy could be replaced by immunotherapy. Therefore, immunotherapy could become a cornerstone of cancer treatment due to its high target specificity.
One might expect enormous benefits from immunotherapy if it could treat cancer at its early stages. However, several daunting outcomes in cancer diagnosis and therapy raise a crucial question, have we been diagnosing cancer timely and accurately? The analysis of tissue-based histopathology and image-based CT-scan/MRI approaches currently count on a human-driven diagnosis. Unfortunately, when it comes to research, a human’s brain has limits, confounded mainly by training and years of experience. Moreover, human experience with cancer diagnosis and treatment has been prolonged despite technological advancements in the past few decades.
Artificial Intelligence
“Artificial Intelligence (AI)” has emerged as a groundbreaking approach in the health care system. AI is the use of computers to model some or all aspects of human intelligence in solving a given problem. Deep learning (DL) consists of several machine learning algorithms among all AI technologies. It is also called deep neural networks (DNNs) e.g. convolution neural network (CNN). These deep networks have evolved during the last few decades. They have changed the way of processing data, images, texts, and speech. Precisely, machine learning applies statistical methods to training data to automatically adjust the parameters of a model, rather than a programmer needing to set them manually. Furthermore, these algorithms run on graphical processing units (GPUs). Therefore, GPUs have the enormous computational power required to train DNNs in a reasonable time.
In 2019, several sessions at the American Association for Cancer Research (AACR) conference in Atlanta, GA, featured the application of AI in improved cancer diagnosis. DL is used to extract information from a patient’s data. It is primarily from radiology images acquired via MRI and CT scans. Images and computers have been used in radiology. However, the development of “The Picture Archiving and Communication System (PACS) has led to the fully approved digital workflow. For example, in a histology approach, images acquired via H&E staining on glass slides are analyzed after the tumor tissue biopsy.
A significant advancement in digital pathology has enabled pathologists to scan whole slides and store pathology images for analysis. So far, there are no set guidelines and data standards. But AI has significantly encroached cancer diagnosis field. Moreover, unlike some people believe, AI will only take some future jobs in Radiology and Pathology. However, AI will complement human knowledge and intelligence in solving problems, at least in cancer diagnosis.
Immunotherapy Weds Artificial Intelligence
Based on several successful data, ongoing immunotherapeutic clinical trials in cancer, and the emergence of critical algorithms in AI, it is not an exaggeration to say “Immunotherapy weds AI” could change the whole understanding of cancer diagnosis and therapy. Time will tell, but the scientific community is buzzing that doctors may not need a biopsy to confirm cancer. Instead, a few medical images (CT scan and/or H&E) and AI could be used to study immunotherapy-treated tumors in patients. At the same time, image-based biomarkers, such as increased CD8 T cell and dendritic cell infiltration, decreased T regulatory cell, decreased M2 polarized macrophages, and decreased myeloid-derived suppressor cells in the tumor could predict which patients respond to cutting-edge immunotherapy, are needed as a confirmation.
Artificial Intelligence Helps in Solving the Cancer-associated Problems
Currently, the scientific community is looking forward to how AI could be instrumental in solving cancer related problems. Particularly, the following cancer-associated problems in a more specific and integrated manner:
(1) How to minimize the toxic side effects of immunotherapy and other standard therapies by precise selection of the patient based on genomic profiling.
(2) One of the major problems in cancer treatment is therapy resistance, where a tumor shrinks after the therapy and comes back strongly and metastasizes to distant organs.
(3) Capturing and blocking key alternative mechanisms of tumor growth such as vasculogenic mimicry, where tumor cells itself functional vessels to outgrow in the hypoxic tumor microenvironment.
(4) Identification of sustainable therapeutic combinations to reduce tumor growth and improve survival.
(5) Advancements in AI algorithms to detect early tumor growth in primary organs and metastatic organs to guide early interventions and treatments.
(6) How AI is going to deal with cancer disparity outcomes in the United States and other parts of the world.
Final Thoughts
As the AI field progresses, we may see parallel advancements in explainability, robustness, and patient data security. However, the funding and expenses associated with several new projects are enormous. Indeed, major funding agencies like the National Institutes of Health and the Department of Defense would play a primary role in the United States. Additionally, a newer initiative like “Global Oncology” would have an essential contribution to advocating this growing AI field in cancer care. Global Oncology could provide effective cancer diagnosis and immunotherapies in all world populations in a strategic way.
Further Readings:
AI applications are yet to explore in full swing even though AI Research is advancing in the healthcare practice. Read AI studies in The Application of Artificial Intelligence (AI) in Medicine and The Scope of Artificial Intelligence in Clinical Oncology
Author:
Dr. B.R. Achyut is a Biologist, Educator, Innovator, Scientific Writer, and Aspiring Entrepreneur in the field of Cancer. He is working as an Associate Scientist at Winship Cancer Institute of Emory University. He has earned a Ph.D. in Cancer Inflammation and Genetics from India at Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow. In the year 2009, he came to the National Cancer Institute at National Institutes of Health, Bethesda for Postdoctoral Fellowship. Later, he moved to Henry Ford Health System, Detroit (2013) and Augusta University, Augusta (2014).
He significantly contributed to the tumor microenvironment area of cancer focusing on Upper GI, breast, and glioblastoma preclinical models. His seminal research work is published in several international cancer and therapy journals. He is recognized as a “Top 10 in 10” list of “Young Professional to Watch 2017” by Augusta Magazine and Augusta Chamber of Commerce. Later, Georgia Trend Magazine in 2017 recognized him as a “40 under 40” in the state of Georgia. He is a big fan of Cricket game. In the past, he established “Augusta Cricket League” and recently, “Suwanee Avengers” Cricket team in Georgia.