Abstract: Cancer is not an unpredictable, shape-shifting disease against which all available weapons must be thrown, but more accurately a population of cells, a “rogue species”, seeking to survive in the ecosystem of the body and following the same principles of evolution as every other living thing. The means by which cancer cell populations develop, and ultimately survive the therapies aimed at killing them are now being understood, genomically mapped and predicted by the principles of evolutionary biology. Yet, this understanding has not translated into the cancer clinic as an enlightened evolutionary strategy of cancer therapy with the goal of cure. Is cure, or “extinction” of the cancer species, in any individual possible? Potential clinical strategies to approach this challenge will be discussed. 

Abstract: The National Cancer Act of 1971 will mark its 50th anniversary in December of 2021. A lot has happened in the last 50 years to unravel some of the mysteries that surrounded cancer in 1971. Progress is especially noteworthy in defining the “omics” of cancer which has enabled defining cancer subtypes based on using whole genome sequencing, RNA-Seq, epigenomic and other advanced molecular technologies to create molecular profiles of individual cancers. Data from these studies and additional patient data from sources such as digital devices and real world evidence, that were not considered significant or didn’t exist in 1971, has led to an unprecedented data “tsunami” in cancer research and care. Most of these data remain to be analyzed. Interestingly, beyond advances in technologies to interrogate cancer cells, the current data surge is driven in large part by the realization that we need data that reflects what is happening in individual patients in real time. This era of what we now refer to as “precision cancer medicine” represents an evolution in thinking about cancer that puts the patient (the phenotype) at the center of cancer research; and achieving broad success in precision cancer medicine is driving a need to understand how cancer evolves in the treatment setting.

Cancer is a complex adaptive system (CAS), where aberrant information (in the information theory sense) expressed across scales (molecular to organism) leads to several emergent properties. Evolution and robustness, two of these emergent properties, makes cancer very difficult to treat, and for metastatic cancer, nearly impossible to cure. Evolutionary pressure that result from nearly all cancer therapies results in resistance to current treatments that is an inevitable and predictable outcome. Solving this problem and improving the success of cancer treatment will depend to a major extent on understanding the effect of these evolutionary and ecological forces in terms of measurable changes (evolutionary biomarkers) in patients. Measurable (perhaps predictive) biomarkers of evolvability (clonal diversity, mutation rates, clonal expansion rates, etc.), represent examples of potential classes of biomarkers that hold promise for new strategies to drive decision making in selecting, monitoring and changing cancer therapies. New ideas and innovative approaches to discover and develop evolutionary biomarkers in various treatment settings requires new ideas and application of advances from the “omics” revolution and associated advanced technologies. Moreover, the development of innovative clinical trial models that enable collecting longitudinal samples and data to support the discovery and validation of evolutionary biomarkers is a critical next step in providing a foundation for the field.

The ISPY 2 Trial for high risk breast cancer is an adaptive ( Bayesian statistics driven) platform clinical trial that represents an excellent supportive model for providing the continuity and learning systems needed to both validate known evolutionary biomarkers and discover new ones. A bit more than 10 years old, ISPY 2 has demonstrated that multiple drugs and combinations can be tested rapidly and at significantly less costs. In doing so, the ISPY 2 trial design has allowed the network of ISPY centers to develop high value longitudinal data and samples that could be transformative to the field of evolutionary biomarkers. The presentation will touch on potential “best in class” biomarkers, the continuum for their development and the role of adaptive platform trials as a supportive model for the field.

In summary, given the promise of evolutionary biomarkers to inform new strategies for drug discovery, risk stratification and prognosis, and their potential to predict therapeutics response and combat multi-drug resistance (MDR), their discovery, development and commercialization is one of the highest value propositions in cancer medicine today.

Abstract: Introduction: Extracellular vesicles (EVs) are nano-sized, membrane enclosed vesicles that are released by cells. While initially thought to be detritus or involved in eliminating waste from cells, a landmark 1996 study demonstrated the functional capabilities of EVs in intercellular communication. Raposo et al. documented that B cell derived EVs could stimulate antigen specific immune responses in recipient T cells. Since then EVs have been recognised as important mediators of intercellular communication by transferring their bioactive cargoes to recipient cells. In this way, they are important physiological mediators in health and disease. The relevance of EVs in understanding biology and their potential as novel therapeutic and diagnostic tools has become clear, as evidenced by the large numbers of research groups studying EVs and the emergence and success of numerous companies. Additionally, the field has established its own thriving society; The International Society for Extracellular Vesicles (ISEV), and associated high impact journal; the Journal of Extracellular Vesicles (JEV).

Notably, over the last two decades, a substantial research effort has been undertaken to understand the role of EVs in cancer. It is now understood that tumour-derived EVs can transfer their contents to influence metastatic behaviour, as well as establish favourable microenvironments and pre-metastatic niches that support cancer development and progression. Furthermore, research has suggested that tumour cells with an increased propensity to adhere secrete more EVs in comparison to tumour cells that take longer to adhere and spread. This could be the domain of metastatic cells which have the capacity to rapidly adhere to various substrata and in this way gain a foothold for growth advantage.

However, notable heterogeneity exists among all secreted EVs due to variations in cellular cargo, cell type, epigenetic factors and cellular differentiation, making the study of EV biology much more complex due to the diverse array of biological effects they can exert on recipient cells. In this study we identified and characterised subpopulations of SKOV3 ovarian cancer cell derived EVs, assessed their role in cellular adhesion and highlight the significance of EV heterogeneity in this context.

Method: SKOV3 ovarian cancer EVs were purified from adherent cell culture by size exclusion chromatography (SEC). Small EVs were then further fractionated into four EV subpopulation pools based on the chromatographic profile of an extended SEC protocol. Separate cell culture surfaces were respectively coated with EVs from each pool and SKOV3 cell adhesion was assessed. EVs in each pool were characterised by nanoparticle tracking analysis, western blotting, MACS Plex Exosomes Flow Cytometry, Nanoview ExoView R100, peptide mass fingerprints via MALDI-TOF mass spectrometry and super-resolution structural illumination microscopy.

Results: Differential SKOV3 cell adhesion was observed to surfaces coated with different EV pools. When blocking CD29, SKOV3 cell adhesion was inhibited suggesting involvement of EV subpopulations in SKOV3 cell adhesion. Characterisation techniques revealed each of the EV pools to be phenotypically distinct and thusly a heterogeneous population of sub-150nm particles. Additionally, a high, yet fluctuating expression of cell adhesion proteins across each of the EV pools, including CD29 and other proteins previously shown to be involved in metastasis was observed. CD29 and the expression of other adhesion proteins correlated with the differential cell adhesion observed. However, CD29 was also shown to be completely anti-colocalised with the most abundantly present EV tetraspanin CD81.

Conclusion: Taken together the data highlights the significance of EV heterogeneity, and the existence of functionally distinct subpopulations of extracellular vesicles. With regard to cancer development and progression, cells may dictate the release of certain subpopulations of vesicles that are involved in these processes as opposed to the EV populations as a whole. Although further experiments remain to be performed to dissect roles, adhesion could be a very rapid way for EVs to contribute to cancer metastasis providing a foothold to migrating cells and offering growth advantages to adhered cells. This will of course be the result of an evolutionary process that has occurred gradually over time to enhance the growth and survival of cancer, and indeed this evolution may be as a result of EV mediated communication. Further studies are needed to investigate whether the EV subpopulations and their unique cargo result from distinct cellular biogenesis pathways or whether this is a stochastic process.

Abstract: The history of modern humans has consisted of a number of demographic events, including population structure, admixture and migration, which have shaped patterns of genetic variation in contemporary populations. In addition, novel genetic and phenotypic adaptations have also evolved in populations in response to dramatic variation in climate, diet, and/or exposure to infectious disease. In the end, present-day patterns of variation in human genomes are a product of both demographic and selective events. One of the “grand challenges” in genomics research is to better understand the connections between evolutionary history, genomic variation and complex traits, including disease susceptibility (e.g. cancers). Indeed, the characterization of genomic diversity, together with phenotype data, is informative for identifying variants that play a role in human adaptation and complex traits. This talk will explore the use of evolutionary and computational approaches to identify polymorphisms that contribute to the development of disease in human populations.

Abstract: We have developed methods using human transcription factor expression libraries to accelerate some developmental processes from 300 days to 4 days and have used these semi-synthetic tissues to accurately evaluate late onset (70 year) diseases like Alzheimer's and extendable to proliferative pathologies including cancer.  We have easily engineerable proliferative systems with doubling times of only 15 minutes that we would like to adapt to display analogous developmental processes.  We have accelerated evolution way of a combination of gigabase-scale computer-controlled DNA synthesis plus machine learning.  We have used this to design and test millions of viral capsids for improved AAV gene therapies.

Abstract: Biology has scorned teleology for 120 years. But engineering has embraced it since the 1940s. All cells exhibit agency and seem to evolve purposefully. This is why cancer has outwitted doctors for 100 years and why stage 3-4 patients are no better in 2020 than in 1930. Evolution was presumed to be random and purposeless, yet cells possess cognition [REF Shapiro all cells are cognitive]. Cell behavior is normally algorithmic, but uniquely responds to novel situations. This is what makes evolution (and cancer) possible.

The origin of biological information is an unsolved cognition problem. Here we propose a new framework by modeling the cell as a Volitional Turing Machine. We model the agency of the cell as a computer that can choose “1” or “0” before writing its next output. This choice is a non-deterministic action of a free agent with sensory capacity and memory. It is not computable from prior states. As well as reading and reacting to its environment, it anticipates future threats, chooses goals and reasons inductively. Computers do none of these things. We replace Aristotle’s four causes (material, formal, efficient, and final) with three orthogonal components: Chemicals, Code and Cognition. The model is consistent with information theory and known limitations of AI, unlike the standard model of evolution. The Volitional Turing Machine encompasses Schrödinger’s “negative entropy” and Davies’ “Demon in the machine.” It explains the futility of reducing cancer to a single random or algorithmic component, and exposes the challenge of understanding this formidable autopoietic evolving life form.

Abstract: Introduction

A major theme of this symposium is that there are strong parallels between the evolutionary origin of species within populations and new concepts for the origin of cancers within the tissues of the body. The analogy is that cancers can be regarded as a new species developing within the host organism.

My presentation will show that to understand this parallel we need to turn the current popular theory of evolution, the Modern Synthesis (Huxley, 1942), on its head. That theory, also called neo-darwinism, is based on the idea that inherited variations are purely random. The better survival of fitter organisms is then attributed to blind (undirected) natural selection. I will show that the reverse is true. Cells and organisms control the stochasticity within them and do so in ways that favour their development and evolution. In doing so, they give directionality to what happens. The process is not completely blind. Many of the physiological control processes by which they achieve this are now fairly well understood.

Cell Selection in development

A theory of Darwinian selection amongst cells in developing tissues was first developed by Kupiec (1983, reprinted 2020), although it had a precursor in Weismann’s idea of selection amongst germ-line cells (Weismann, 1896). Kupiec (2009) has also emphasised the role of stochasticity in cellular development.

Writing a commentary on the 2020 republication of Kupiec’s 1983 paper, Andras Paldi (2020) has succinctly expressed the parallel:

“The apparently predetermined gene expression patterns that characterise the defined cell phenotypes are the results of a selective stabilization of some patterns through interactions between the cells. This way of framing one of the modern biology’s central questions calls for the same reasoning Charles Darwin proposed to explain the evolution of biological species. The idea that spontaneous variation followed by selective stabilization of some of these variants can account for the emergence of new cell types during ontogenesis places the evolution of the species and cell types on the same theoretical ground.”

I have emphasised the phrase relevant to this symposium. It speaks for itself.

Harnessing of chance

To these ideas I will add three further concepts:

1. Stochasticity becomes functional when it is harnessed. This idea was first presented at the “New Trends in Evolution” meeting of The Royal Society and The British Academy in 2016 (Noble 2017). There are many ways in which such harnessing occurs in organisms.

(a) DNA replication. The natural error rate in DNA replication is very high. In a genome of 3 billion base pairs nearly a million ‘chance’ errors occur. The final error rate when the genome is passed on to daughter cells is so small that hardly a single error is passed on. That high accuracy is achieved by the cell itself, which processes the raw error rate through a 3-stage process to achieve faithful transmission of hard genetic information. This is one of the reasons why DNA is dead outside an organism. Contrary to Schrödinger’s (1942) idea of the genetic molecule being an aperiodic crystal, DNA cannot self-replicate as a crystal does. It needs to be pampered by the cell to maintain its meaning and utility (Noble, 2018).

(b) Developing new antibodies. Cells also have the means to produce many new variant DNA sequences in a highly targeted region of the genome coding for the variable part of an immunoglobulin (Shapiro, 2012a,b).

(c) Developing anticipatory behaviour. A theory of neural selection was first developed by Edelman (1978). My brother and I have developed this idea to show how, by analogy with the harnessing of chance in the immune system, organisms can generate new forms of anticipatory behaviour (Noble & Noble, 2017, 2018, 2020).

2. Variation is not then random and can be at least partially directed (Noble & Noble, 2017).

This fact is ubiquitous in biology. Just consider what happens to the energy of photons randomly poured out into space by the sun. The few that hit the earth might encounter a green plant. The plant harnesses that energy through photosynthesis. The outcome, the growth of the plant and subsequently the growth of animals as they consume plants, is very far from random. That process, like the harnessing of randomness in the immune system, is unconscious. But, around the time of the Cambrian explosion (Ginsburg & Jablonka, 2019), the harnessing of random processes at all levels in organisms became an active, often conscious, process. Organisms that possess conscious agency then take part in social selection, originally identified by Darwin (1871) in the example of sexual selection.

There can therefore be two forms of directed evolution, unconscious and conscious. The form relevant to cancer is unconscious, but it is still directed.

3. Cells communicate their DNA, RNA and regulatory states. They can do this via intercellular exchange of information within extracellular vesicles, EVs (Edelstein et al 2019). This process includes the transmission of soma information to the germ-line and can therefore form a mechanism for the inheritance of acquired characteristics. This is almost the same theory that Darwin (1871) developed as the theory of gemmules. The EVs, I suggest, contain Darwin’s postulated gemmules (Noble, 2019).

The implications for the development of cancers are important. Those implications are the subject of another presentation in this symposium (Bonner, Willms & Noble, 2020).

Abstract: Breast cancer is the most commonly diagnosed cancer and it is the second leading cause of cancer-related death in women worldwide¹. More than 90% of breast cancer deaths are due to metastasis^2,3. Although the exact number is unknown, estimates suggest in the US in 2020 over 168,292 women are living with metastatic breast cancer (mBC)⁴. Three out of every four of these women were initially diagnosed with Stage I-III disease who later progressed⁴. Despite advances in prevention and therapeutics, treatments to eradicate mBC do not exist. The median survival of patients with mBC is 3 years, with less than 22% of patients surviving beyond 5 years⁴.  By 2050, the global incidence of female breast cancer is predicted to reach 3.2 million new cases per year⁵.  As the number of cases increases, so will the cost to treat breast cancer. In 2018, the national cost to treat all cancer was $150.8B, with female breast cancer leading all cancer sites at an estimated $19.7B⁷.

The overwhelming majority of breast cancers (75%) are deemed hormone receptor positive (HR+) either expressing the estrogen receptor or progesterone receptor⁸.  The backbone of HR+ breast cancer treatment remains endocrine therapy (ET)¹. ET includes selective estrogen receptor modulators (SERMS: tamoxifen), aromatase inhibitors (AIs: letrozole, analstrozle, exemestane) and selective estrogen receptor down regulators (SERDs: fulvestrant). In the metastatic setting, ET is often supplemented with additional targeted therapies such as CDK4/6 inhibitors (palbociclib, ribociclib, abemaciclib), PI3K inhibitors (alpelisib) and/or mTOR inhibitors (evorlimus)¹. There is no standard consensus for the optimal sequencing of these agents.

Resistance (intrinsic and acquired) are major issues with all known targeted therapies. The differences between intrinsic and acquired resistance are not clear, and each may operate through different mechanisms. Patient heterogeneity, tumor microenvironment, crosstalk between signaling pathways, the microbiome, environmental factors and more, contribute to resistance. Each therapy or class of therapies is also likely to have distinct mechanisms of resistance. Being able to detect intrinsic and/or acquired resistance to alter a patient’s therapy remains a long-standing goal of precision medicine (PM).

Biomarkers are the cornerstone of PM, enabling clinicians to use the altered expression of DNA, RNA, proteins, metabolites or combinations thereof to direct a clinical action. “Biomarkers of response” (BoR) are a subset of biomarkers that are capable of identifying patients that will respond to specific therapies and those that will not. Genomic-based BoR efforts have had limited success. This is not surprising as additional factors such as age, diet, environment, microbiome and other factors influence drug response. At present there are few, if any, BoRs available for clinicians to monitor drug response and resistance in mBC.

Recent advancements in metabolomics offer new promise to identify BoRs with bona fide clinical utility. Metabolism is influenced by both genetics and the environment. Through metabolomics, differences in metabolism between healthy and diseased states can be uncovered. Furthermore, by measuring metabolite changes, it is feasible to construct a metabolic fingerprint differentiating drug response and drug resistance.  Here we will describe metabolomic approaches that have identified BoRs with high predictive accuracy to differentiate mBC patients that did and did not respond to CDK4/6 inhibitors. Upon validation in larger cohorts, the results of these studies could lead to a paradigm shift in medicine, moving away from stepwise medicine to one that is optimized for individual success. This has the potential to dramatically improve outcomes, reduce unnecessary exposure to side effects and diminish the financial toxicity.

1. Ballinger, T. J., Meier, J. B. & Jansen, V. M. Frontiers in Oncologyvol. 8 308 (2018). 2. Chaffer, C. L. & Weinberg, R. A. Sciencevol. 331 1559–1564 (2011). 3.The war on cancer Departments of Pharmacology and Medicine, Dartmouth Medical School, Hanover, NH 03755, USA (Prof M B Sporn MD).4.Mariotto, A. B., Etzioni, R., Hurlbert, M., Penberthy, L. & Mayer, M. Cancer Epidemiol. Biomarkers Prev.26, 809–815 (2017).5.Tao, Z. Q. et al. Cell Biochem. Biophys.72, 333–338 (2015).

Abstract: Three key questions set the context for daring to believe: Why is cancer so hard, How are we doing in the fight against it, and Is a cure realistic?

1. Cancer is hard because it is a natural consequence of evolution.  Cancer cells do what all cells are supposed to do, mutate, evolve, and increase in fitness compared to their neighbors to outlive them. To survive, cancer cells create a cancer friendly ecosystem to hijack our immune systems and suppress an immune response such that cancer is less a disease and more a phenomenon, the result of a basic compromise that is fundamental to evolution.
2. How are we doing against this goliath? Following the advent of chemotherapy, there was a dearth of progress until 20 years ago with the introduction of targeted therapies. Over the past 10 years, the concept that the key to fighting cancer may already be inside each of us has led to extraordinary progress in the field of immuno-oncology.  Yet we are far from declaring victory.  Cancer cells mount their own Darwinian fight to survive against the body’s immune attack and, despite success with therapeutic approaches that re-activate T cells, only a small percentage of patients respond and most of those that do respond, ultimately relapse. It has become increasingly clear that turning on the immune system isn’t enough; the key is to make sure it doesn’t get turned off AND that it is not over-activated. 
3. Against these challenges, is a cure realistic?  There is still much to be learned about how to best leverage our very dynamic immune systems in fighting cancer, balancing adaptive and innate immunity in killing cancer cells without triggering auto-immune reactions.  The tremendous amount of translational analysis that is incorporated into clinical trials today makes it increasingly possible to identify those patients who benefit from immune activating therapies as well as those who are less susceptible to undesirable autoimmunity.  These insights are an essential component of current clinical trial design to maximize the benefit: risk ratio in the treatment of cancer patients. AND these insights will be key to taking us to the next level: identifying patients at risk for cancer and developing cures!

Abstract: In patients, de novo lethal cancer evolves capacities for both metastasis and resistance. Therefore, cancers in different patients appear to follow the same eco-evolutionary path that independently manifests in affected patients. This convergent outcome, that always includes the ability to metastasize and exhibit resistance, demands an explanation beyond the slow and steady accrual of stochastic mutations. We suggest that cancer starts as a speciation event when a unicellular protist breaks away from its multicellular host and initiates a cancer clade within the patient. As the cancer cells speciate and diversify further, some evolve the capacity to evolve: evolvability. By generating and maintaining considerable heritable variation, the cancer clade can, with high certainty, serendipitously produce cells resistant to therapy and cells capable of metastasizing. Understanding that cancer cells can swiftly evolve responses to novel and varied stressors create opportunities for adaptive therapy, double-bind therapies, and extinction therapies; all involving strategic decision making that steers and anticipates the convergent coevolutionary responses of the cancers