Evolutionary scalpels for dissecting tumor ecosystems

https://doi.org/10.1016/j.bbcan.2016.11.005Get rights and content

Abstract

Amidst the growing literature on cancer genomics and intratumor heterogeneity, essential principles in evolutionary biology recur time and time again. Here we use these principles to guide the reader through major advances in cancer research, highlighting issues of “hit hard, hit early” treatment strategies, drug resistance, and metastasis. We distinguish between two frameworks for understanding heterogeneous tumors, both of which can inform treatment strategies: (1) The tumor as diverse ecosystem, a Darwinian population of sometimes-competing, sometimes-cooperating cells; (2) The tumor as tightly integrated, self-regulating organ, which may hijack developmental signals to restore functional heterogeneity after treatment. While the first framework dominates literature on cancer evolution, the second framework enjoys support as well. Throughout this review, we illustrate how mathematical models inform understanding of tumor progression and treatment outcomes. Connecting models to genomic data faces computational and technical hurdles, but high-throughput single-cell technologies show promise to clear these hurdles. This article is part of a Special Issue entitled: Evolutionary principles - heterogeneity in cancer?, edited by Dr. Robert A. Gatenby.

Introduction

Medical science has long known tumors to consist of multiple different parts [1]. Before the advent of modern genome sequencing, researchers routinely observed that different parts of a tumor could vary in cell morphology, karyotype, immunogenicity, and drug sensitivity [2], reported in frequent publications about tumor heterogeneity [3]. More recently, dramatic advances in genomics, transcriptomics, and technologies to isolate and characterize single cells have enabled new quantitative understandings of heterogeneity. Large-scale studies have revealed both that each tumor bears a diverse complement of somatic mutations, some occurring at very low frequencies (intratumor heterogeneity), and that the genomic make-up of tumors can vary in clinically important ways across patients suffering from the same cancer type (interpatient heterogeneity) [4], [5].

The field of precision medicine attempts to address interpatient heterogeneity by identifying drug targets specific to each patient, based on personalized molecular information. Yet technologies for precision medicine make abundantly clear that tumors are genetically and epigenetically complex entities that cannot be reduced to a mere list of mutations. Observations both in vivo and in vitro reveal that tumors are dynamic, heterogeneous populations of cells, a symbiotic biological community, which may respond in unexpected ways to therapy [6], [7], [8], [9], [10]. If therapy is to effectively take personalized genomic information into account, it must be informed by evolutionary principles. In this review, we will explore these principles and connect them to clinical findings.

Many previous reviews have discussed how heterogeneous tumors evolve [2], [3], [11], [12], [13], [14], [15], [16], highlighting in particular the complexity of tumor ecosystems [17], [18], the relationship between tumor heterogeneity and progression to metastasis [19], [20], [21], and the importance of considering tumor evolution when designing therapy [6], [7], [8], [9], [10], [21]. Many computational methods have been designed to grapple with the onslaught of data produced by genomic [22] and epigenomic [23] studies. New mathematical models are being invented to interpret the output of these computational methods, to understand the evolutionary processes underlying tumor progression and response to treatment [24], [25]. Amidst the growing literature, essential principles in evolutionary biology recur time and time again. This review offers an exposition of these principles, highlighting clinical scenarios. Our perspective is shaped deeply by mathematical and computational methods, our “scalpels” for dissecting the evolutionary history of heterogeneous tumors. In this review, we present the cutting edge of these scalpels without algebraic notation.

We begin by confronting the problem of drug resistance (2 Intrinsic versus acquired drug resistance, 3 Combination therapy: can evolutionary principles offer a new hope?), reviewing the arguments, modeling frameworks, and clinical findings that have contributed to current understanding of evolution of tumors to evade targeted therapy and progress to metastasis. We will describe a predominant school of thought, based on population genetic modeling, that champions hard-hitting administration of multiple therapies early in disease progression to prevent the evolution of resistance [24], [26], [27]. This strategy has seen outstanding success in some areas, particularly in transforming childhood acute lymphoblastic leukemia from an immediately fatal diagnosis to one in which most patients are cured [28]. Some research, however, counsels caution in the enthusiastic extrapolation of this principle to all cancer types, suggesting that sometimes the best we may hope for is long-term restriction of a tumor to a small, nonthreatening volume [7], [29].

As we proceed, we will distinguish between two different concepts of intratumor heterogeneity: the tumor as an ecosystem (Section 4), versus the tumor as an organ (Section 5). While both of these terms have been used before to describe aspects of tumor biology [8], [30], they have not been contrasted explicitly. These two concepts involve different assumptions about how portions of a tumor may regrow following reduction via effective therapy. Considering the tumor as an ecosystem, regrowth of a tumor is an instance of the evolutionary phenomenon of adaptive radiation [31], [32], in which mutant cells manage to explore and exploit underused ecological niches, in which they can proliferate. Considering the tumor instead as an organ, regrowth of a tumor is akin to anatomic regeneration or morphallaxis [33], [34], in which cells are directed by biochemical signals to restore lost function. This distinction corresponds to a key difference long understood in evolutionary biology – adaptation by selection of inherited alterations that arose randomly [35], [36], versus selection of intrinsic biological programs capable of producing or restoring a desired phenotype [37].

Throughout this review, we will use quantitative concepts that are the bread-and-butter of evolutionary biology – population sizes, fitness, mutation frequencies, and measures of diversity. Sometimes, modeling approaches treat these concepts as given, as if they were easily readable from the book of nature. In truth they are complex, have definitions that change from author to author and over time, and are difficult to measure. These challenges have motivated many computational and experimental advances (Section 6 and Box 1). We close with three open questions provoked by the research reviewed (Section 7).

Section snippets

Intrinsic versus acquired drug resistance

Poor prognosis of many cancer types most often results from the failure of therapy to clear a treatment-resistant portion of the tumor. Resistance is often classified as intrinsic or acquired, depending on whether the cancer fails to respond initially to treatment (intrinsic resistance), or recurs as a resistant tumor following initial response to therapy (acquired resistance). Genomic mechanisms conferring resistance to targeted therapy include amplification of genes targeted for inhibition or

Combination therapy: can evolutionary principles offer a new hope?

The extreme genetic diversity present in a large tumor means that it is nearly inevitable for some cells to harbor alterations conferring resistance to therapy [52], [53]. In glioblastoma (GBM), genomic amplifications commonly occur in multiple signaling pathways, suggesting that therapeutic targeting of only one pathway will merely allow growth of cellular populations that exploit an alternate pathway [54]. In colorectal cancers initially sensitive to anti-EGFR antibody therapy, rapid

Adaptive radiation and ecological niches: understanding the cancer ecosystem

Interactions between cancer cells extend far beyond the mechanisms of drug resistance, and investigation has revealed a rich cancer ecosystem. A beautiful study of evolution of heterogeneous tumors 35 years ago demonstrates how the concept of an adaptive radiation, while originally used to describe the origin of animal species specializing in different niches, applies to cancer ecosystems as well [91]. In melanoma tumor cells, metastatic capability was measured by number of lung metastases

Morphallaxis and self-regulation: is the tumor an organ?

The studies discussed above explore how different populations of cells interact, together determining the fate of the tumor. The fitness – or growth potential – of each type of cell depends on which other cell types are present in their local environment, a scenario that fits comfortably within frameworks established by theoretical evolutionary biologists to describe frequency-dependent fitness and spatially heterogeneous environments. There is a deeper sense, however, in which the fates of

How can we measure intratumor heterogeneity “in the wild”?

So far, we have discussed experimental and theoretical approaches for understanding tumor evolution, progression, and drug resistance. We have taken as given researchers' ability to identify and quantify specific tumor subclones. In clinical settings, identifying genomic subclones involves statistical methods and substantial guesswork. It is therefore difficult to apply many population genetic approaches directly, as they presuppose the ability to genetically sequence individual organisms and

Conclusions

Principles and models in evolutionary biology can help to unify and synthesize the avalanche of studies investigating intra-tumor heterogeneity. Our review leaves open three broad questions:

1) Is “hit hard, hit early” with combination drug therapy a feasible solution for cancers in general, or are there fundamental limits to this strategy, related to a tumor's propensity to restore heterogeneity following environmental insult? Elegant and general mathematical models, experiments investigating

Definitions (sidebar or box with caption: “concepts in cancer evolution”)

    Adaptive radiation

    Diversification of an evolving population into multiple lineages, each adapted to a separate ecological niche [31], [32].

    Morphallaxis

    Anatomical regeneration following an amputation, relying on cellular redifferentiation and migration to restore lost organism function [33], [34].

    Secondary mutation

    A somatic mutation that typically is observed only after one or more specific initial mutations. The appearance and growth of the initial mutation may establish conditions that favor

Conflicts of interest

The authors declare no conflicts of interest.

Transparency document

Transparency document.

Acknowledgements

D.I.S.R. is supported by the US National Library of Medicine (T15 LM007079). P.G.C. thanks Arnold Levine, Suzanne Christen, and the Institute for Advanced Study for their hospitality. R.R. acknowledges funding from the NIH (U54 CA193313, R01 CA185486, and R01 CA179044).

References (196)

  • J. Stelling et al.

    Robustness of cellular functions

    Cell

    (2004)
  • D.E. Goldberg et al.

    Outwitting evolution: fighting drug-resistant TB, malaria, and HIV

    Cell

    (2012)
  • J.J. Westendorf et al.

    Expression and function of Fas (APO-1/CD95) in patient myeloma cells and myeloma cell lines

    Blood

    (1995)
  • Z.-W. Li et al.

    Tumor microenvironment and drug resistance in hematologic malignancies

    Blood Rev.

    (2006)
  • V.S. Jones

    Cytokines in cancer drug resistance: cues to new therapeutic strategies

    Biochim. Biophys. Acta

    (2016)
  • D. Berner et al.

    The genomics of organismal diversification illuminated by adaptive radiations

    Trends Genet.

    (2015)
  • J. Calbo

    A functional role for tumor cell heterogeneity in a mouse model of small cell lung cancer

    Cancer Cell

    (2011)
  • J.W. Olejarz et al.

    Evolution of worker policing

    J. Theor. Biol.

    (2016)
  • D.R. Welch

    Tumor heterogeneity—a ‘contemporary concept’ founded on historical insights and predictions

    Cancer Res.

    (2016)
  • G.H. Heppner et al.

    Tumor heterogeneity: biological implications and therapeutic consequences

    Cancer Metastasis Rev.

    (1983)
  • G.H. Heppner

    Tumor heterogeneity

    Cancer Res.

    (1984)
  • International Cancer Genome Consortium

    International network of cancer genome projects

    Nature

    (2010)
  • B. Vogelstein

    Cancer genome landscapes

    Science

    (2013)
  • R.A. Gatenby et al.

    Lessons from applied ecology: cancer control using an evolutionary double bind

    Cancer Res.

    (2009)
  • R.A. Gatenby

    A change of strategy in the war on cancer

    Nature

    (2009)
  • M. Greaves et al.

    Clonal evolution in cancer

    Nature

    (2012)
  • S. Aparicio et al.

    The implications of clonal genome evolution for cancer medicine

    N. Engl. J. Med.

    (2013)
  • C. Willyard

    Cancer therapy: an evolved approach

    Nature

    (2016)
  • L.M.F. Merlo et al.

    Cancer as an evolutionary and ecological process

    Nat. Rev. Cancer

    (2006)
  • C. Swanton

    Intratumor heterogeneity: evolution through space and time

    Cancer Res.

    (2012)
  • A. Marusyk et al.

    Intra-tumour heterogeneity: a looking glass for cancer?

    Nat. Rev. Cancer

    (2012)
  • R.A. Burrell et al.

    The causes and consequences of genetic heterogeneity in cancer evolution

    Nature

    (2013)
  • N. McGranahan et al.

    Biological and therapeutic impact of intratumor heterogeneity in cancer evolution

    Cancer Cell

    (2015)
  • G.H. Heppner et al.

    Tumor heterogeneity is fundamental to the tumor ecosystem

    Oncology (Williston Park, N.Y.)

    (2014)
  • D.P. Tabassum et al.

    Tumorigenesis: it takes a village

    Nat. Rev. Cancer

    (2015)
  • F.R. Miller et al.

    Cellular interactions in metastasis

    Cancer Metastasis Rev.

    (1990)
  • D. Neelakantan et al.

    Intratumoral heterogeneity: clonal cooperation in epithelial-to-mesenchymal transition and metastasis

    Cell Adhes. Migr.

    (2015)
  • H. Zhou et al.

    Clonal cooperativity in heterogenous cancers

    Semin. Cell Dev. Biol.

    (2016)
  • N. Beerenwinkel et al.

    Cancer evolution: mathematical models and computational inference

    Syst. Biol.

    (2015)
  • O. Stegle et al.

    Computational and analytical challenges in single-cell transcriptomics

    Nat. Rev. Genet.

    (2015)
  • P.M. Altrock et al.

    The mathematics of cancer: integrating quantitative models

    Nat. Rev. Cancer

    (2015)
  • J.H. Goldie et al.

    Rationale for the use of alternating non-cross-resistant chemotherapy

    Cancer Treat. Rep.

    (1982)
  • F. Grossi

    Sequential, alternating, and maintenance/consolidation chemotherapy in advanced non-small cell lung cancer: a review of the literature

    Oncologist

    (2007)
  • T.L. Cheng

    Seven great achievements in pediatric research in the past 40 y

    Pediatr. Res.

    (2016)
  • R.A. Gatenby et al.

    Adaptive therapy

    Cancer Res.

    (2009)
  • J.B. Losos

    Adaptive radiation, ecological opportunity, and evolutionary determinism

    Am. Nat.

    (2010)
  • J.B. Losos et al.

    Adaptive radiation: the interaction of ecological opportunity, adaptation, and speciation

  • K. Agata et al.

    Unifying principles of regeneration I: epimorphosis versus morphallaxis

    Develop. Growth Differ.

    (2007)
  • S.E. Luria et al.

    Mutations of bacteria from virus sensitivity to virus resistance

    Genetics

    (1943)
  • R.E. Lenski et al.

    Mutation and selection in bacterial populations: alternatives to the hypothesis of directed mutation

    Proc. Natl. Acad. Sci. U. S. A.

    (1989)
  • Cited by (0)

    This article is part of a Special Issue entitled: Evolutionary principles - heterogeneity in cancer?, edited by Dr. Robert A. Gatenby.

    View full text