Role of intratumoural heterogeneity in cancer drug

Authors: Nicholas A Saunders Fiona Simpson Erik W Thompson Michelle M Hill Liliana Endo‐Munoz Graham Leggatt Rodney F Minchin Alexander Guminski

Publish Date: 2012/08/03

Volume: 4, Issue: 8, Pages: 675-684

PDF Link

Abstract

Drug resistance continues to be a major barrier to the delivery of curative therapies in cancer Historically drug resistance has been associated with over‐expression of drug transporters changes in drug kinetics or amplification of drug targets However the emergence of resistance in patients treated with new‐targeted therapies has provided new insight into the complexities underlying cancer drug resistance Recent data now implicate intratumoural heterogeneity as a major driver of drug resistance Single cell sequencing studies that identified multiple genetically distinct variants within human tumours clearly demonstrate the heterogeneous nature of human tumours The major contributors to intratumoural heterogeneity are i genetic variation ii stochastic processes iii the microenvironment and iv cell and tissue plasticity Each of these factors impacts on drug sensitivity To deliver curative therapies to patients modification of current therapeutic strategies to include methods that estimate intratumoural heterogeneity and plasticity will be essentialCure or control of disseminated disease remains the greatest challenge facing cancer clinicians/scientists and the greatest cause of patient mortality Advances in surgery and radiation oncology provide cures for many forms of malignancy However these advances are unlikely to produce substantial improvements in cures for patients with metastatic disease The control of malignancies beyond the primary site of the tumour requires systemic therapeutic strategies Systemic chemotherapy‐based treatments for cancer emerged in the 1940s to 1960s amid considerable resistance from the clinical community DeVita Chu 2008 Single drug treatments for cancer were widely practiced until the 1960s when DeVita and coworkers championed the concept of combination chemotherapy The rationale for their approach was to combine agents with different modes of action thereby increasing the likelihood of synergistic anti‐cancer effects Devita Schein 1973 By the mid‐1970s combination chemotherapy had increased the complete remission rate for Hodgkins lymphoma from 20 to 80 and for lymphosarcoma from 15 to over 50 Devita Schein 1973 With few exceptions combination chemotherapy is now standard practice when treating both primary and secondary tumours Despite these advances a significant fraction of advanced human malignancies remain refractory to curative attempts with conventional chemotherapeutics mainly due to inherent or acquired drug resistanceIn the last decade there has been a large effort to identify specific mutations within tumours that could be exploited as therapeutic targets We now have clinical experience with several new classes of ‘targeted’ and ‘non‐targeted’ therapies such as anti‐angiogenic drugs anti‐stromal drugs immune modulators epigenetic modifiers and inhibitors of various growth factors and their signalling pathways Patient response to these drugs has varied from profound curative responses Kwak et al 2010 Rosti et al 2012 through to transient Flaherty et al 2010 Sampson et al 2010 or poor responses Garraway Janne 2012 Despite the success of targeted and non‐targeted approaches to treating cancers the emergence of resistant disease continues to be a significant cause of patient mortalityTumour Drug resistance can be inherent or acquired and is mediated by multiple biochemical processes operating individually or in combination Fodale et al 2011 Known factors that lead to drug resistance include i induction of drug transporters ii activation of DNA repair iii changes in drug metabolism iv gene amplification or mutation of target proteins and v changes in survival/apoptotic pathways The ATP‐binding cassette ABC family of drug transporters comprises 48 genes that code for transmembrane pumps that are selectively involved in the efflux of small molecule drugs and toxins Fukuda Schuetz 2012 Some of these transporters have been shown to play a specific role in pumping cytotoxic drugs out of the cell preventing the accumulation of cytotoxic concentrations within the cell and hence invoking drug‐resistance Fukuda Schuetz 2012 DNA repair pathways comprise a complex network of proteins able to sense DNA damage eg ATM ATR Chk1/2 BRCA1 or p53 through to the machinery required to repair the damage Drug metabolizing enzymes such as the cytochrome P450 family of enzymes or the glucuronyl transferases are responsible for the biotransformation of many anti‐cancer drugs and their activity contributes to the modulation of intracellular drug levels This contributes to the sensitivity of cancer cells to cytotoxic drugs In addition to these regulators of drug sensitivity gene amplification of receptors such as EGF receptor targeted by tyrosine kinase inhibitors can often compromise the efficacy of therapies Finally the relative activities of pro‐apoptotic and anti‐apoptotic survival pathways contribute to the sensitivity of a cancer cell to a cytotoxic stimulus see Chonghaile Letai 2008 Engelman 2009 Combined these broad overlapping mechanisms are the main biochemical determinants of cancer cell sensitivity to cytotoxic drugs excluding anatomical or diffusional considerations Therapeutic strategies to modify drug transporters see Haar et al 2012 drug metabolism and survival pathways see Chonghaile Letai 2008 Engelman 2009 have all been developed and trialed in patients However further resistance frequently occurs followed by disease relapse and progression Thus knowing the main biochemical contributors to resistance has not led to the development of tests that are predictive of tumour behaviour nor has it led to substantive improvements in patient outcomesThe emergence of resistance to the diverse range of drugs available for cancer treatment is indicative of the dynamic nature of tumour tissue Recent evidence emerging from studies in which the tumour environment was interrogated suggests that a more fundamental driver of resistance is intratumoural heterogeneity Ding et al 2012 Navin et al 2011 Ruiz et al 2011 Xu et al 2012 These studies highlight the likelihood that tumours comprise cancer cells that vary in their sensitivity to chemotherapeutics due to genotypic or phenotypic variation This is an important finding because the basis of resistance has significant implications for the management of cancer patients Heterogeneity in this context refers to variation in tissue response tissue composition tissue physiology tissue phenotype and tissue genotype For example genomic heterogeneity may arise through heritable genetic and epigenetic mechanisms and is exemplified by the presence of discrete clonal variants within a tumour Navin et al 2011 the potential presence of tumour initiating subpopulations of cells Bonnet Dick 1997 Lapidot et al 1994 or cells with a characteristic ‘mutator’ phenotype see Kolodner et al 2011 Phenotypic heterogeneity may result from genomic heterogeneity but also can result from stroma/tumour cell interactions tumour cell/tumour cell interactions or simply as a result of the stochastic nature of biological processes Fig 1 All of these contributors to intratumoural heterogeneity are likely to be operative simultaneously and highlight why resistance is likely to be dynamicScheme depicting the basis for heterogeneity in drug responses within a tumour The different clonal variants present within tumours are represented by the different coloured cells Heterogeneity within a clonal variant due to stromal interaction is marked by whilst heterogeneity attributable to plasticity/EMT is marked by ⁁⁁ Finally heterogeneity in an identical clone due to stochastic variation is markedUnderstanding the biological and genetic basis for how cells acquire heterogeneity has important implications for how we manage patients The acquisition of intratumoural heterogeneity is frequently modelled on evolutionary principles Gerlinger et al 2012 Polyak 2007 In these models a tumour is assumed to derive from a single founder cell that has acquired a mutation in a critical gene This mutation is passed on to progeny that are subject to further lesional events resulting in the production of progeny that continue to acquire genetic/epigenetic mutations leading to a fully transformed malignancy comprising many clonal variants Ding et al 2012 Gerlinger et al 2012 Polyak 2007 Two models of tumour evolution have been proposed the cancer stem cell model and the clonal evolution model Polyak 2007 Shackleton et al 2009 Both accommodate the generation and expansion of genotypic and phenotypic variants within tumours The major point of difference between the models is whether tumour initiating and self‐renewal activity is restricted to a fixed subpopulation of cells or is shared by all the variant clones within the tumourThe cancer stem cell model proposes that there is a fixed rare subpopulation of cancer cells that possess stem‐like activity with respect to their self‐renewal capacity and ability to initiate a tumour in xenotransplant models Bonnet Dick 1997 Lapidot et al 1994 Shackleton et al 2009 This model is strongly supported by data from human leukaemias such as acute myeloid leukaemia AML Bonnet Dick 1997 Lapidot et al 1994 Cells enriched from patients with AML could be divided into populations that differed in the expression of specific surface molecules The bulk of AML cells was CD32−CD38− and when injected into immunocompromised mice could not initiate a tumour By contrast the rare CD32+CD38− fraction of cells could initiate tumours with high efficiency Bonnet Dick 1997 Lapidot et al 1994 There have now been a number of studies in other tumour types including pancreatic cancer breast cancer head and neck cancer and medulloblastoma which have also reported the presence of rare stem‐like cells within tumours Al‐Hajj et al 2003 Bao et al 2006 Li et al 2007 Prince et al 2007 In contrast the clonal evolution model allows most if not all cells to retain a capacity for self‐renewal and tumour initiating activity Thus from a therapeutic point of view the cancer stem cell model would require ablation of the cancer stem cell population to invoke a cure whereas the clonal evolution model would require the ablation of all clonal variants to invoke a cure However both models allow for the generation of genotypically discrete clonal variants that could differ with respect to chemotherapeutic sensitivity Recent reports have indicated that tumour cell subpopulations can trans‐differentiate into one another indicating that the rigid requirement of the cancer stem cell model that tumour initiating activity is restricted to a fixed population of tumour cells may be questioned Chaffer et al 2011 Gupta et al 2011 Roesch et al 2010 Whilst these studies have only been reported for breast cancer and melanoma they do provide an important conceptual framework to unify the two models

Keywords: