Tumor Dormancy and MYC Inactivation: Pushing Cancer to the

Brink of Normalcy

Catherine M. Shachaf and Dean W. Felsher

Division of Oncology, Departments of Medicine and Pathology Stanford University, Stanford, California

Abstract

Upon MYC inactivation, tumors undergo proliferative arrest,cellular differentiation, apoptosis and in some cases, perma-nently revoke tumorigenesis. In liver tumor cells, we recentlyshowed that MYC inactivation uncovers stem cell propertiesand triggers differentiation, but in this case, their neoplasticproperties are restorable by MYC reactivation. Thus, whereasoncogene inactivation can push cancer to the brink ofnormalcy, some cells retain the latent capacity to turncancerous again, arguing that they may exist in a state oftumor dormancy. (Cancer Res 2005; 65(11): 4471-4)

Introduction

Cancer is one of the most common causes of death. Since theidentification of the first activated proto-oncogene, the targetedinactivation or repair of these mutant gene products has beenstudied as a potentially therapeutically useful approach for thetreatment of neoplasia (1). To date, only a few drugs have beenidentified that have been successfully shown to have clinicalefficacy through the inactivation of a specific oncogene (2, 3).However, recent work in experimental transgenic model systemssuggests that the targeted inactivation of a single oncogene maybe generally effective in the treatment of some cancers (4).Transgenic mouse models provide a powerful approach to

preclinically evaluate the mechanisms by which genetic eventsinitiate and sustain tumorigenesis. The development of conditionaltransgenic model systems has provided the unique ability todirectly interrogate when oncogene inactivation will be effective ininducing tumor regression (4, 5). A common general feature ofoncogene inactivation in a wide spectrum of tumors arising indifferent cellular lineages is that upon oncogene inactivationtumors undergo proliferative arrest, differentiation, and/or apop-tosis. Thus, oncogene inactivation in some cases seems able toreverse the neoplastic phenotype.Surprisingly, under some circumstances, even the brief

inactivation of a single oncogene can be sufficient to inducesustained tumor regression, strongly indicating that, at least insome cases, oncogene inactivation can cause permanent loss ofthe neoplastic phenotype, indicating that cancer has been‘‘revoked’’ (6). These observations are consistent with clinicalexperience in human cancer patients, where disease can be curedin some cases by administration of a limited number of cycles ofchemotherapy, radiation therapy, or other biological therapies.The observation that such treatments can ever be efficacioussuggests that a therapeutic approach that transiently affects the

biology of a cancer cell can be fully effective. In many cases,therapy results in the irreversible loss of a cancer phenotypethrough necrosis or apoptosis, but in other cases, it is clear thatthe cancer cells are still present and either that they no longerbehave as cancer cells or that they can remain dormant for longperiods of time.Recent results from several laboratories suggest that MYC

inactivation in some cancers will not revoke tumorigenesis;rather, it will induce a transient loss of neoplastic character,marked by normal phenotype, that can be fully restored by MYCreactivation (6–9). To develop successful new treatments forcancer, it is important to learn to distinguish when oncogeneinactivation induces stable tumor regression associated with apermanent loss of a malignant phenotype, versus when oncogeneactivation merely results in a reversible state of tumordormancy.

MYC Inactivation: Tumorigenesis Revoked versusTumor Dormancy

The consequences of MYC inactivation and reactivation seem todepend upon the particular type of cancer. In MYC-inducedlymphoma, MYC inactivation results in differentiation, robustapoptosis, and complete tumor regression (10). These tumorsregress despite the complexity of the genomic events acquired bythe tumors (11). Thus, MYC inactivation in lymphoma commonlyresults in a permanent loss of a neoplastic phenotype. In contrast,in MYC-induced osteogenic sarcoma, MYC inactivation inducesdifferentiation and dramatically reduces the rate of cellularproliferation but does not induce significant apoptosis. In thisin vivo setting, MYC inactivation seems to lead to the return ofnormal bone morphology function (6). Reactivation of MYC inthese apparently differentiated tumor cells either has no conse-quence or induces apoptosis; only very rare cells restore theoperation of neoplastic pathways. These studies show that MYCinactivation is able to both reverse and revoke a neoplasticphenotype.In contrast, in MYC-induced hepatocellular carcinoma, MYC

inactivation initially induces differentiation and only later isfollowed by gradual apoptosis of the majority of the tumor cells.The differentiation process seems to uncover stem cell propertiesbecause after MYC inactivation, the cells are able to differentiateinto multiple cell lineages that form structures characteristic ofnormal hepatocytes and biliary cells (7). However, in this setting,reactivation of MYC causes these differentiated tumor cells torapidly become tumorigenic. We have suggested the followingmodel to interpret these results (Fig. 1). MYC inactivationuncovers the stem cell properties of liver tumor cells, permittingthem to differentiate into normal liver cell lineages; however,some of these normal-appearing cells actually remain in a latentstate, such that upon MYC reactivation they rapidly restore theirneoplastic features. As described below, we propose that this

Requests for reprints: Dean W. Felsher, Division of Oncology, Departments ofMedicine and Oncology, CCSR 1150, MC 5151 Stanford University, Stanford, CA, 94305.Phone: 650-498-5269; Fax: 650-725-1420; E-mail: dfelsher@stanford.edu.

I2005 American Association for Cancer Research.

www.aacrjournals.org 4471 Cancer Res 2005; 65: (11). June 1, 2005

Review

Cancer Research. on September 23, 2020. © 2005 American Association forcancerres.aacrjournals.org Downloaded from

happens because some of the differentiated tumor cells arecancer stem cells that can regain their neoplastic properties.In this fashion, MYC inactivation induces a state of tumordormancy.We performed a variety of experiments to rule out the

possibility that tumors that reemerged after MYC reactivationwere derived from hepatocytes that were initially only partiallymalignantly transformed. We found that normal hepatocytes fromthe MYC transgenic mouse model used could not persist whentransplanted into new hosts, even when coinoculated with tumorcells. We also did this test using liver tumor cells from a lungmetastasis, obtaining identical results, arguing against the remotepossibility that normal hepatocytes that may be present in aprimary tumor had metastasized along with cancer cells to thelung. Finally, using array comparative genomic hybridization, weconfirmed that the parental tumor, transplanted tumor, dormanttumor, and restored tumor contained an identical geneticsignature, providing very strong evidence that they were derivedfrom the same clone.Another consideration was whether the liver tumors that

recurred upon MYC reactivation were derived from normal-appearing differentiated tumor cells or from rare cancer cells thatwere undetectable by histologic analysis. These rare cells, whereasquiescent, may have failed to undergo differentiation because theyhad acquired some additional genetic hits that allowed them tobecome independent of MYC. Two lines of evidence argued againstthis explanation for tumor recurrence. First, using biolumines-cence imaging, we were able to do highly quantitative measure-ments on the rate of tumor regression and reemergence. Basedupon the signal intensity, it seemed that a large fraction of thetumor cells were proliferatively expanding upon MYC reactivation.Second, in the tumors that returned upon MYC reactivation, a

second round of MYC inactivation induced tumor regressionidentically to the parent tumor. This observation argued againstthe likelihood that secondary genetic events acquired during initialtumor formation were responsible for compromising sensitivity toMYC inactivation.Several other recently published results have also described that

oncogene inactivation can induce tumor regression and thatoncogene reactivation leads to rapid tumor reformation. Thus,MYC-induced skin and pancreatic islet cell tumors undergoregression upon MYC inactivation and that tumors reform quicklyafter MYC reactivation (8). Similarly, MYC-induced breast cancerundergoes regression upon MYC inactivation that is fully revokedby MYC reactivation (12). In RAS-induced melanoma, oncogeneinactivation similarly induces tumor regression; however, in thiscase, residual microscopically detectable cancers remained thatrapidly reemerged as large tumors upon RAS reactivation (13).Similarly, in Gli-induced skin tumors, residual disease remainingafter oncogene inactivation reemerged rapidly upon Gli reactiva-tion (14). Thus, in many cases, oncogene inactivation does notresult in complete and sustained tumor regression (with the caveatthat in some cases there might have been a failure to achieve fullinactivation of the tumor-driving oncogene).Taken together, MYC transgenic mouse studies suggest that

there are circumstances when inactivation of the MYC oncogenecan induce a state of tumor dormancy. These results have pot-ential implications for the treatment of cancer discussed furtherbelow.

Tumor Dormancy and Cancer Stem Cells

Tumor dormancy, observed clinically, is generally thought toreflect the existence of residual tumor cells that fail to respond to a

Figure 1. MYC inactivation uncovers stem cell–like properties in liver cancer that can induce a state of tumor dormancy. MYC-induced liver cancers retain pluripotentcell properties that are revealed upon MYC inactivation, with tumor cells differentiating into both hepatocytes and bile duct cells. Some cells may remain as dormantcancer cells that are histologically normal, which upon MYC reactivation emerge from their dormant state and regain their tumorigenic properties.

Cancer Research

Cancer Res 2005; 65: (11). June 1, 2005 4472 www.aacrjournals.org

Cancer Research. on September 23, 2020. © 2005 American Association forcancerres.aacrjournals.org Downloaded from

conventional treatment regimen (15–18). Residual tumor cells maybe kept in check by various mechanisms, including the hostimmune system, tumor cell microenvironment, and/or angiogenicor autocrine growth factor capability (19–23). Our findings are notexclusive of these possibilities. The notion that cancer can be‘‘reprogrammed’’ is not new and has been recently validatedthrough Herculean efforts that illustrate how normal mice can begenerated, albeit rarely, via nuclear transfer from a cancer cell(24, 25).How might oncogene inactivation allow tumor cells to uncover

latent stem cell–like properties? One possibility is that tumors,even if they seem morphologically or histologically homogenous,consist of different cell populations with discrete differences atthe molecular level. This is consistent with the evidence fromother groups indicating that cancer might originate from thetransformation of a cancer stem cell. In leukemia and breastcancer, a specific subpopulation of the tumor is able to propagatethe tumorigenic properties of the cancer. If liver cancer inducedby MYC activation arises specifically from stem cells, then uponMYC inactivation one might predict that these stem cells wouldregain the ability to resume their physiologic program and toundergo differentiation into progenitor cells as well as moremature cell lineages. However, because of their stem cell roots,these differentiated cells might be capable of maintaining apopulation of cells that can self-renew at a slower rate. In thesecells, which retain some stem cell features, reactivation of theMYC oncogene could result in the restoration of the neoplasticprogram.Still, other possibilities can be envisaged. For example, MYC

overexpression may simply usurp or defeat certain normalprograms of cell fate. MYC plays important roles in embryonicand organ development; thus, it is conceivable that the programsaffected by MYC are shared with programs found in stem cells.Upon MYC inactivation, these programs would be released fromthe domination of MYC, such that the normal program canresume, similar to what occurs during normal organ developmentor regeneration in the stem cell compartment. Two otherpossibilities are that different cell populations are transformedby MYC, such that MYC inactivation releases the differentpopulations to their normal cellular programs, resulting indifferent cell lineages; or that a normal cell fuses with a tumorcell in the primary tumor, somehow leading the heterokaryon toexhibit stem cell properties. These possibilities seem to offer poorexplanations of the response of hepatocytes to cycles of MYCactivation and inactivation, because the molecular mapping of thetumors showed that the tumors are clonal even when MYC hasbeen inactivated and reactivated for several rounds, includingfollowing several series of transplantation.

Clinical and Therapeutic Implications of TumorDormancy

Our results suggest that targeted therapy may induce tumorregression and an apparent clinical elimination of the malig-nancy by histologic criteria. There may be instances when whathas actually occurred is restoration of normal tissues thatinclude a population of occult and dormant tumor cells. Thus, adetermination that a cancer has been truly eliminated maysometimes require more than a measurement of the size of atumor or an examination of the tissue histology to rule out thepresence of dormant tumor cells that nevertheless appear andbehave normally.The notion that cancer can become dormant upon treatment is

supported by decades of clinical experience (26). Treatment ofmany types of cancers, including breast cancer, lymphoma,leukemia, and sarcomas, can be associated with long-term reg-ression of tumors, only to be followed by relapse at a much latertime (15–18). The specific type of cancer can often predict theoutcome of treatment. Thus, hematopoietic cancers are perhapsone of the best examples of how treatment outcome is influencedby the specific type of tumor. High-grade lymphomas like Burkitt’slymphoma and diffuse large cell lymphoma frequently respond tochemotherapy with rapid apoptosis. In these cases, cancer isreadily cured. In contrast, low-grade lymphomas, such as follicularlymphoma, whereas also responding to therapy, are rarely seen toregress completely (27).The recent findings of transgenic mouse studies may offer

explanations of what has been seen for many years in the clinic.Some types of cancer exhibit sustained regression to treatment andremain in remission. Others will undergo initial regression butremain present in an occult dormant state that reemerges uponcessation of therapy, leading to a rapid reoccurrence. Clinicaloutcome will no doubt reflect the biology unique to specificsubtypes of cancer and their underlying features. A successfultherapeutic strategy for cancer will therefore require an under-standing of the nature of how genetic events and epigenetic contextof a tumor can be used to anticipate consequences of gene productinactivation, such that one can achieve a complete elimination oftumor cells or a permanent reversal of tumorigenesis.

Acknowledgments

Received 4/5/2005; accepted 4/18/2005.Grant support: National Cancer Institute (D.W. Felsher), American Society of

Clinical Oncology Young Investigator Award, UCSF Liver Center pilot feasibility grant,Stanford University Digestive Disease Center pilot award (D.W. Felsher), WeilandFamily fellowship (C.M. Shachaf), and Flight Attendant Medical Research InstituteYoung Scientist Clinical Award (C.M. Shachaf).

We thank the members of the Felsher laboratory for their suggestions.

Tumor Dormancy and MYC Inactivation

www.aacrjournals.org 4473 Cancer Res 2005; 65: (11). June 1, 2005

References

1. Bishop JM. Molecular themes in oncogenesis. Cell1991;64:235–48.

2. Druker BJ, Sawyers CL, Kantarjian H, et al. Activity of aspecific inhibitor of the BCR-ABL tyrosine kinase in theblast crisis of chronic myeloid leukemia and acutelymphoblastic leukemia with the Philadelphia chromo-some. N Engl J Med 2001;344:1038–42.

3. Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safetyof a specific inhibitor of the BCR-ABL tyrosine kinasein chronic myeloid leukemia. N Engl J Med 2001;344:1031–7.

4. Felsher DW. Cancer revoked: oncogenes as therapeutictargets. Nat Rev Cancer 2003;3:375–80.

5. Giuriato S, Rabin K, Fan AC, Shachaf CM, Felsher DW.Conditional animal models: a strategy to define whenoncogenes will be effective targets to treat cancer.Semin Cancer Biol 2004;14:3–11.

6. Jain M, Arvanitis C, Chu K, et al. Sustained loss of aneoplastic phenotype by brief inactivation of MYC.Science 2002;297:102–4.

7. Shachaf CM, Kopelman AM, Arvanitis C, et al. MYCinactivation uncovers pluripotent differentiation andtumour dormancy in hepatocellular cancer. Nature2004;431:1112–7.

8. Pelengaris S, Abouna S, Cheung L, Ifandi V, ZervouS, Khan M. Brief inactivation of c-Myc is notsufficient for sustained regression of c-Myc-inducedtumours of pancreatic islets and skin epidermis. BMCBiol 2004;2:26.

9. Flores I, Murphy DJ, Swigart LB, Knies U, Evan GI.Defining the temporal requirements for Myc in theprogression and maintenance of skin neoplasia. Onco-gene 2004;23:5923–30.

10. Felsher DW, Bishop JM. Reversible tumorigenesis byMYC in hematopoietic lineages. Mol Cell 1999;4:199–207.

11. Karlsson A, Giuriato S, Tang F, Fung-Weier J, Levan G,Felsher DW. Genomically complex lymphomas undergo

Cancer Research. on September 23, 2020. © 2005 American Association forcancerres.aacrjournals.org Downloaded from

Cancer Research

Cancer Res 2005; 65: (11). June 1, 2005 4474 www.aacrjournals.org

sustained tumor regression upon MYC inactivationunless they acquire novel chromosomal translocations.Blood 2003;101:2797–803.

12. Boxer RB, Jang JW, Sintasath L, Chodosh LA. Lack ofsustained regression of c-MYC-induced mammaryadenocarcinomas following brief or prolonged MYCinactivation. Cancer Cell 2004;6:577–86.

13. Chin L, Tam A, Pomerantz J, et al. Essential role foroncogenic Ras in tumour maintenance. Nature 1999;400:468–72.

14. Hutchin ME, Kariapper MS, Grachtchouk M, et al.Sustained Hedgehog signaling is required for basal cellcarcinoma proliferation and survival: conditional skintumorigenesis recapitulates the hair growth cycle.Genes Dev 2005;19:214–23.

15. Pantel K, Otte M. Identification and characterisationof minimal residual disease in solid tumors. Acta MedAustriaca 2000;Suppl 52:8–12.

16. Stock W, Estrov Z. Studies of minimal residual

disease in acute lymphocytic leukemia. Hematol OncolClin North Am 2000;14:1289–305, viii-ix.

17. Demicheli R. Tumour dormancy: findings andhypotheses from clinical research on breast cancer.Semin Cancer Biol 2001;11:297–306.

18. Kwok PC, Lam TW, Lam PW, et al. Randomizedcontrolled trial to compare the dose of adjuvantchemotherapy after curative resection of hepatocellularcarcinoma. J Gastroenterol Hepatol 2003;18:450–5.

19. Schirrmacher V. T-cell immunity in the inductionand maintenance of a tumour dormant state. SeminCancer Biol 2001;11:285–95.

20. Uhr JW, Marches R. Dormancy in a model ofmurine B cell lymphoma. Semin Cancer Biol 2001;11:277–83.

21. Delsanto PP, Romano A, Scalerandi M, PescarmonaGP. Analysis of a ‘‘phase transition’’ from tumor growthto latency. Phys Rev E Stat Phys Plasmas Fluids RelatInterdiscip Topics 2000;62:2547–54.

22. Folkman J. Role of angiogenesis in tumor growth andmetastasis. Semin Oncol 2002;29:15–8.

23. Korah R, Boots M, Wieder R. Integrin a5h1promotes survival of growth-arrested breast cancercells: an in vitro paradigm for breast cancerdormancy in bone marrow. Cancer Res 2004;64:4514–22.

24. Blelloch RH, Hochedlinger K, Yamada Y, et al. Nuclearcloning of embryonal carcinoma cells. Proc Natl AcadSci U S A 2004;101:13985–90.

25. Hochedlinger K, Rideout WM, Kyba M, Daley GQ,Blelloch R, Jaenisch R. Nuclear transplantation, embry-onic stem cells and the potential for cell therapy.Hematol J 2004;5 Suppl 3:S114–7.

26. Hochedlinger K, Blelloch R, Brennan C, et al.Reprogramming of a melanoma genome by nucleartransplantation. Genes Dev 2004;18:1875–85.

27. Archuleta TD, Armitage JO. Advances in follicularlymphoma. Semin Oncol 2004;31:66–71.

Cancer Research. on September 23, 2020. © 2005 American Association forcancerres.aacrjournals.org Downloaded from

2005;65:4471-4474. Cancer Res   Catherine M. Shachaf and Dean W. Felsher  the Brink of NormalcyTumor Dormancy and MYC Inactivation: Pushing Cancer to

  Updated version

  http://cancerres.aacrjournals.org/content/65/11/4471

Access the most recent version of this article at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/65/11/4471.full#ref-list-1

This article cites 23 articles, 5 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/65/11/4471.full#related-urls

This article has been cited by 10 HighWire-hosted articles. Access the articles at:

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/65/11/4471To request permission to re-use all or part of this article, use this link

Cancer Research. on September 23, 2020. © 2005 American Association forcancerres.aacrjournals.org Downloaded from