We must find ways to detect cancer much earlier

“Everyone here has the sense that right now is one of those moments when we are influencing the future.” —Steve Jobs

Every year, cancer kills approximately 10 million people worldwide. Of those who die, two thirds do so because they were diagnosed with advanced disease. A new paradigm in the approach to cancer is overdue. Covid-19 has already altered conversations and expectations within the medical community and is forcing a rethinking of many public health issues.

To contemplate a transformative approach for the postpandemic cancer landscape, The Oncology Think Tank (TOTT) was created in June 2020, bringing together a diverse group of thought leaders, researchers and oncologists from academia and industry. Meetings were held remotely, at least once a week and sometimes twice weekly for four months. The burden of TOTT was to formulate a fresh, compassionate, patient-centric, effective and radically different vision for health care’s approach to cancer. This opinion paper will focus on what TOTT believes is the best way forward with a goal of reducing the number of patients who are diagnosed with, or develop, advanced stage cancers and die.

There has been universal agreement that the best way to abolish cancer’s terrible impact on the world is through early detection. In October, Cancer Research UK (CRUK) published an article in the Lancet Oncology laying out an important cross-sectoral vision for a future where no cancer will be detected too late to treat. TOTT proposes not just early stage I and II cancer detection as understood today but to reach farther back in time to spot the earliest detectable precancerous perturbations.

This approach would move the research focus from studying established cancers to identifying the earliest signs of conversion from wellness to a malignant state, essentially, by studying the health trajectory of each individual during three states: wellness, wellness-to-disease transition and, finally, disease. Knowledge gained by such an approach would provide the opportunity to develop strategies to prevent the third stage of managing the clinical diagnosis of the disease. And by definition, it should be easier and more successful because an exponential increase in the complexity, as reflected by disease-perturbed networks, is expected during cancer evolution and progression.

Realization of such a vision will require pragmatic shifts in our understanding of the biology of early cancer, but also shifts in our attitudes and investments. For example, the current practice of periodic screening will have to be replaced by a more dynamic format of real-time screening to understand the biological changes involved in the wellness-to-disease transition.

The notion of a change from screening for individual cancers in many people to screening every person for many cancers will need to be accepted as a fundamental new strategy. Ultimately, the job of the oncologist of the future will be to treat and prevent the emergence of disease as opposed to treating established disease. TOTT acknowledges the challenge of how best to effect preventive measures once the earliest signs of cancer are detected, but the group does strongly underscore the importance of personalizing treatment using this approach of precision medicine.

To meet the demands of this new model, there need to be scientific efforts, public policy changes and broad education efforts occurring in parallel. It is equally vital that the existing reimbursement system be expanded to support the proposed paradigm changes by adopting a patient-centric, population-based approach to health care.

The salient recommendations of TOTT can be divided into three main areas:

Research to solve the technical challenges of the earliest cancer detection: Detecting the earliest cancer changes is, first and foremost, a technologic challenge. Developing biological insights into the dynamics of wellness-to-illness transition demands a complex systems approach. Systems thinking rests on the principle that biological information is quantized thereby restricting the number of possible values or states. One can argue that the richest source of easily measurable biologic quanta in the human body is blood. Along with breath, saliva, sweat, tears, urine and stool, blood carries molecules, vesicles and circulating tumor cells shed from relevant tissues that may contain critical transition signals.

Moreover, blood has innumerable undiscovered biomarkers. Emerging, sophisticated technologies to study epigenetic, genetic, transcriptomic, metabolomic and proteomic biomarkers from multiple compartments, along with studies of the microbiome, can identify organ-specific blood proteins that serve as proxies for their cognate networks to determine precisely when they become cancer-perturbed for every individual. In addition, we recommend that efforts be made to develop more effective approaches to the earliest possible detection of stage I and II cancers.

Third-generation DNA sequencing technology is capable of performing whole genome sequences and transcriptome analyses rapidly while revealing epigenetic modifications. Moreover, multimodality approaches, for example, combining mass spectrometry with scanning and imaging devices, help distinguish between benign versus malignant states. There are sophisticated measurements available to identify minimal residual disease already, and these now must be employed to detect the earliest possible transitions and minimal initial disease. Finding rare circulating tumor cells at either stage is a good example. Wearable sensors can automatically collect dynamic information in real time longitudinally. Current health watches can monitor vital signs, oxygen saturation, glycemic indices, exercise patterns and symptoms experienced, with emerging continuous sensors targeting molecular analytes. Through artificial intelligence and machine learning trained on billions of data points collected for each individual, the danger signs of impending disease could be recorded far ahead of its actual clinical appearance.

TOTT acknowledges that cancer screening for the earliest cancer transitions, precancerous/early stage I and II emerging cancer detection, disease appearance, progression, therapy resistance, metastasis and recurrence needs to be viewed as an ever-changing, time-dependent and longitudinal patient problem. Diagnosing or screening for cancer in one patient (n=1) will have limited impact on the population of people at risk. But studying and following large numbers of individual patients longitudinally using a systems approach will have a chance to significantly affect cancer’s global impact. Setting the right goals and financially incentivizing them will accelerate discovery.

Policy changes to create an ecosystem that encourages innovation and rewards technologies that improve early detection: The Covid-19 pandemic offers a practical example of how to bring about policy changes that seek solutions for the population as a whole rather than for isolated individuals and to accomplish the changes with alacrity. It has never been more clear that the seemingly glacial pace of drug development and the massive bureaucracy obstructing and frustrating cancer research is indefensible. If it is possible to get a trial studying remdesivir in patients with Covid-19 open and to accrue patients in less than 10 days from receipt of the first protocol draft, it makes little sense for a cancer trial to take six to 12 months to open. There needs to be a compelling imperative for pioneering the early prediction and prevention of cancer.

Is it possible that there is little to no sense of urgency in cancer protocols? Creating policy changes that lead to an effective, achievable new future for the cancer paradigm will require a dedicated and focused coalition involving all of the stakeholders including patients, oncology care providers, researchers, the public at large, insurance carriers, the media and policy makers. Practically speaking, the National Cancer Institute (NCI) would need to dedicate resources and focus on early detection (wellness-disease transition). Groups such as the National Comprehensive Cancer Network (NCCN) would need to embrace the charge to develop a stepwise protocol that would screen the earliest cancer transitions, improve our ability to detect precancerous perturbations and employ a systematic approach to reverse the transition. Changes in public policy must be considered to facilitate coverage for this new type of screening.

The current regulatory policy focuses on screening for a single type of cancer using traditional metrics, whereas multicancer early detection tests could require new metrics (e.g., longitudinal blood omics quantification) because they may optimize features such as overall population cancer detection rates, low false-positive rates and high predictive value, rather than high single-cancer sensitivity.

The Food and Drug Administration will need to rapidly develop novel approaches to assessing test performance and benefit-risk assessments for multicancer approaches that use a variety of genomic and comprehensive blood omics recognition technologies that have heretofore not been applied to cancer screening. Moreover, payer policy needs significant modernization. Centers for Medicare and Medicaid Services (CMS) policy precludes coverage and payment for prevention and early detection unless specific legislative exceptions are made. CMS and commercial payers presently cover and reimburse for most technology and treatment in advanced cancer and almost nothing in prevention and early detection. This must change to keep pace with technology that can improve early cancer detection.

Future screening tests must not only identify a cancer at its earliest detectable stage, but also determine whether it is a cancer that is expected to become a clinical problem for the patient. An important task would be to define the series of tests required for validation and confirmation of findings in order to ensure that interventions are warranted. For example, if circulating tumor cells are identified in an otherwise healthy individual, a supplemental CT or PET scan may reveal the precise origin of the cancer. If the tumor is not yet detectable by existing imaging, however, are there other biomarkers that would prove useful to identify the tissue of origin? Single-cell simultaneous sequencing of RNA and DNA is already developed and could detect “cancers of unknown primary” with even greater precision. In what order should these further tests be conducted?

Education to generate a broad coalition with a clear call to action: Success for any such paradigm shift requires meaningful education at every level from patients and families to the treating oncologists to the cancer research communities and ultimately to the public. There must be clear explanations of why a change is urgently needed, as well as models showing how success will dramatically lessen cancer’s impact on the world and how such sweeping programs can democratize health care by removing inequalities. National societies focused on cancer such as the American Society of Clinical Oncology, the American Association for Cancer Research and the American Society of Hematology already have the platforms including annual meetings, publications and outreach programs to facilitate the education of all stakeholders.

Interdisciplinary studies must be implemented, formally combining research in oncology with cutting-edge research in evolutionary biology, physics, computational biology, machine learning and artificial intelligence dedicated to following the natural history starting with wellness to disease to progression. Such studies should not be limited to humans because there are many examples in nature where species are protected from developing cancer (e.g., elephants, blue whales), and a formal discipline dedicated to the comparative study of cancers across species could be critical in developing preventive measures eventually.

One strategy for the implementation of such a broad and far-reaching program would be to embark on the systems approach targeting a group at high risk of developing cancer—notably cancer survivors. Of 1.7 million cancers diagnosed annually in the U.S., one in five arise in a cancer survivor. There are currently 16.9 million cancer survivors, and the number will increase to 26 million by 2040. Practically every major institution caring for cancer patients has a large population of cancer survivors coming regularly for periodic checkups. Resources should be dedicated to obtaining multiple tissues such as blood, urine, saliva and stools periodically from cancer survivors to identify markers for, and develop insights into, their wellness-to-recurrence transitions.

Expanding the scope of large-scale noninvasive projects and encouraging health care systems and academic centers to establish biobanks with multiple specimens collected from every individual who walks through their door, will create a rich resource that can change the current paradigm of cancer health care with deliberate speed.

TOTT acknowledges that a high proportion of cancers arising in survivors will be relapses of already highly complex malignant cell populations, and that these will be fundamentally different from those neoplasms emerging de novo in a stepwise fashion. We also understand that the biomarkers detected in second primary cancers in this high-risk group are unlikely to be the same as those found in first-time cancer patients. But their detection would provide proof of principle for the wellness-to-disease transition model, and add to the knowledge needed for broadening this approach to elevated-genetic-risk segments of the healthy population.

As Thomas Kuhn famously pointed out in his book The Structure of Scientific Revolutions, “Paradigms gain their status because they are more successful than their competitors in solving a few problems that the group of practitioners has come to recognize as acute.” Cancer survivors may provide such a successful opportunity, and hopefully, the results will be lifesaving not just for cancer survivors but for everyone.

The co-authors of this essay are listed below. If not otherwise noted, they have indicated that they have no potential conflicts of interest.

Azra Raza, M.D., is a professor of medicine at the Columbia University Irving Medical Center (CUIMC). She has received grants from GRAIL and Regeneron Pharmaceuticals.

Abdullah M. Ali, Ph.D., is a research scientist at CUIMC. He has no potential conflicts directly related to this work. Outside of it, Ali is currently in receipt of grants from GRAIL, Tolero Pharmaceuticals and Regeneron. He is also a consultant to Vor Biopharma.

Aris Baras, M.D., M.B.A., is senior vice president of Regeneron and founder and general manager of the Regeneron Genetics Center, a wholly owned subsidiary of Regeneron. He is an employee of Regeneron and owns stock in the company.

Robert Gallo, M.D., is a professor of medicine and director of the Institute of Human Virology at the University of Maryland School of Medicine.

Robert A. Gatenby, M.D., is a senior member at the Moffitt Cancer Center.

Anisa Hassan, M.D., H.M.D.C., is a hematologist-oncologist staff physician at Freeman Health System.

Mark L. Heaney, M.D., is an associate professor of medicine at CUIMC.

Joseph G. Jurcic, M.D., is a professor of medicine at CUIMC. He has no potential conflicts directly related to this work. Outside of it, Jurcic receives research funding paid to Columbia University from AbbVie, Arog Pharmaceuticals, Astellas Pharma, Celgene, Daiichi Sankyo, Forma Therapeutics, Genentech, Kura Oncology, PTC Therapeutics and Syros Pharmaceuticals. He has consulted for or serves on advisory boards for AbbVie, Actinium Pharmaceuticals and Novartis.

Stavroula Kousteni, Ph.D., is a professor of physiology and cellular biophysics at CUIMC.

Richard Larson, M.D., is a professor of medicine at the University of Chicago. He has acted as a consultant or adviser to Agios, Amgen, Ariad Pharmaceuticals/Takeda Pharmaceutical Company, Astellas Pharma, Celgene/Bristol-Myers Squibb, CVS Caremark, Epizyme, MorphoSys and Novartis. And he has received clinical research support from Astellas, Celgene, Cellectis, Daiichi Sankyo, Forty Seven, Gilead Sciences, Novartis and Rafael Pharmaceuticals, as well as royalties from UpToDate.

Frank Laukien, Ph.D., is president and CEO of Bruker Corporation, as well as a shareholder of the company. Bruker is a Nasdaq-traded life-science tools company that makes and sells scientific instruments, which can also be used for cancer research, among many other uses and applications. Cancer research tools represent less than 1 percent of Bruker’s revenue.

Steven H. Lin, M.D., Ph.D., is an associate professor of radiation oncology at the University of Texas MD Anderson Cancer Center. He has received grants from BeyondSpring and Hitachi Chemical Diagnostics. And he serves on advisory boards for STCube, BeyondSpring and AstraZeneca.

Guido Marcucci, M.D., is a professor of medicine at City of Hope National Medical Center.

Jayesh Mehta, M.D., is a professor of medicine at Northwestern University’s Feinberg School of Medicine.

Siddhartha Mukherjee, M.D., Ph.D., is an associate professor of medicine at CUIMC.

Joshua Ofman, M.D., is chief medical officer at GRAIL and is an equity owner of the company.

Patrizia Paterlini, M.D., Ph.D., is a professor of oncology at University Paris Descartes and a founder and shareholder of Rarecells; the company is the exclusive licensee of the ISET patents, which Paterlini co-invented.

Kenneth Pienta, M.D., is a professor of urology, oncology, and pharmacology and molecular sciences at the Johns Hopkins School of Medicine. He is a consultant for Cue Biopharma.

Samuel Sia, Ph.D., is a professor of biomedical engineering at Columbia University and co-founder of Rover Diagnostics.

Seema Singhal, M.D., is a professor of medicine at Northwestern University’s Feinberg School of Medicine.

B. Douglas Smith, M.D., is a professor of oncology at the Johns Hopkins School of Medicine.

Patrick Soon-Shiong, M.D., is executive chairman of ImmunityBio; adjunct professor of surgery at the University of California, Los Angeles; visiting professor at Imperial College London; and executive chairman of the Los Angeles Times.

David P. Steensma, M.D., is an associate professor of medicine at the Dana-Farber Cancer Institute. He has no conflict-of-interest disclosures directly related to this work. Steensma acts as a consultant for Pfizer, Cellarity, Taiho, Onconova Therapeutics and Celgene/Bristol-Meyers Squibb.

John Wrangle, M.D., is an associate professor of medicine at the Medical University of South Carolina.

Leroy Hood, M.D., Ph.D., is a professor and the chief strategy officer of the Institute for Systems Biology and senior vice president and chief science officer of Providence St. Joseph Health.