Shaobo:

This is an area of my own research specialization, but given the fine contributions above, I will restrict my attention to two primary themes that should add some further nuance and precision to the discussion: (1) the radical divergence of the new model of heterogeneity from its predecessor; (2) and as always, the clinical relevance of the new model, in particular in the domains of therapeutic response, anti-resistance, and predictive biomarkers, with examples drawn from human clinical data.

The recognition of intratumor heterogeneity for cellular phenotypes is indeed ancient history. In fact it began over 150+years ago: the great Ur-pathologist Rudolf Virchow observed intratumoral pleomorphism of cancer cells, namely morphological heterogeneity of contained malignant cells within individual tumors, in the mid-1800's (circa 1859), followed by demonstrations of the intratumoral functional and genetic heterogeneity, especially the demonstrations over 30 to 40 years ago of distinct subpopulations of cancer cells within tumors, differentiated as to tumorigenicity, therapeutic resistance, and metastatic potential [1,2], including demonstrations by molecular analysis of specifically genetic variability among individual cancer cells [3].

THE NEW HETEROGENEITY MODEL:

The differences however from these historical roots are vast, and new focus however has shifted dramatically and uniquely to:

(1) CELLULAR HETEROGENEITY FOCUS:

Intratumor phenotypic heterogeneity at the single-cell level, aka cancer cell heterogeneity (or cell-intrinsic heterogeneity).

(2) MULTIMODAL HETEROGENEITY:

This is now within the domains of not just tumor biological and histopathological domains, but also in domains of:

- GENETIC HETEROGENEITY (including heterogeneity of normal stem/progenitor cells),

- EPIGENETIC HETEROGENEITY given that epigenetics has been recognized as an important factor in generating non-genetic heterogeneity, an area that has received to date insufficient attention from heterogeneity theorists but stands to be one of the most clinically relevant domains of application [30;13]. From the epigenetic perspective on heterogeneity, combination anticancer therapeutics of cytotoxic or biological agents with epigenetic agents, in particular HDAC and DNMT inhibitors, enables an multiple-angle assault on heterogeneous tumor cell populations that minimizes the evasion opportunities for drug-resistant cells to escape and generate a drug-resistant tumor.

- DETERMINISTIC CELLULAR HETEROGENEITY (multiple fairly stable phenotypic states)

- STOCHASTIC CELLULAR HETEROGENEITY (transient differences in phenotypes between isogenic cells that share the same deterministic phenotypic state),

- CANCER STEM CELL (CSC) HETEROGENEITY, and

- METASTATIC CELLULAR HETEROGENEITY,

- TUMOR MICROENVIRONMENT HETEROGENEITY: there is a new recognition of tumor microenvironment heterogeneity, and with not only intratumor heterogeneity of tumor cells, but also of stromal cells and non-cellular components of microenvironments, and note that the heterogeneity of tumor microenvironments translates into heterogeneous selective pressures experienced by the tumor cells themselves.

(3) CONTINUUM MODEL:

A new "continuum model" of intratumoral heterogeneity in which resident cells reside in different states of stemness or differentiation, on a spectrum of degrees.

(4) CELLULAR PLASTICITY:

This too is a new focus and direction, and represents a wholesale rejection of the classical perspective that cancer stem cell (CSC)-to–non-CSC conversion is a strictly unidirectional process: the plasticity postulate within the cancer cell heterogeneity viewpoint represents as new model of bidirectional interconversions between non-CSCs and CSCs, entailing that non-CSCs can continually create CSC populations throughout tumorigenesis. In this sense, the bold new hypothesis is that the stem cell state is reentrant, with molecular and cellular mechanisms enabling the reentering of the stemness state [27,28], and newly viewing cellular differentiation not as, traditionally, a unidirectional process, but a plastic process where cancer cells can dedifferentiate into more primitive, stem-like phenotypes, these plastic phenotypic shifts finally helping to account the discontinuous behavior of cancer evidenced in some cancers remaining dormant for extended durations (months or years) after therapy, only to relapse later.

THE TUMOR ECOSYSTEM

This gives rise to a model of heterogeneous tumors as complex ecosystems, wherein even a minor tumor subpopulation can influence global tumor growth, that is growth of the entire tumor, actively maintaining tumor heterogeneity and in turn possibly confounding predictive biomarkers and adversely facilitating therapeutic resistance.

CLINICAL RELEVANCE OF THE HETEROGENEITY MODEL:

THERAPEUTIC RESPONSE

One area of critical therapeutic relevance arises from a strategy of heterogeneity modulation, the targeted reduction in tumoral and cellular heterogeneity via heterogeneity-reductive pathways, which is one species of what I have otherwise termed transformational oncology. So as just one example (there are dozens), I will take NY-ESO-1:

NY-ESO-1 is an immunogenic cancer-germline/testis antigen aberrantly expressed in several human malignancies, including epithelial ovarian cancer (EOC) where it is deployed for immunotherapy, and it's known that a frank majority of NY-ESO-1–positive EOC tumors display a heterogeneous expression pattern of this antigen, causing immune and clinical response to NY-ESO-1 peptide vaccine therapy be limited secondary to this heterogeneous expression pattern among (inter-) and within (intra-) tumors [4,5].

Where the clinical therapeutic connection arise is that intratumoral NY-ESO-1 expression is dependent on promoter methylation (both promoter-specific and global DNA methylation status). And as the seminal Roswell Park Cancer Institute study [5] demonstrated, DNA hypomethylation induction with (5-)azacitidine (Vidaza), a DNA methylation inhibitor that is a special form of DNMT (DNA methyltransferase) inhibitor, can restore NY-ESO-1 expression in nonresponder cells. In essence, what we have here is that promoter methylation regulates NY-ESO-1 expression heterogeneity in epithelial ovarian cancer (EOC), and that either treatment with a DNMT inhibitor (azacitidine ) or an HDAC inhibitor (decitabine/Dacogen), functionally restores NY-ESO-1 expression in nonexpressing (nonresponder) EOC cell lines. And I am glad to report that this has recently [this January, 2014] received clinical confirmation in a Phase I trial [10] showing that DNA methyltransferase (DNMT) inhibitors augment NY-ESO-1 vaccine therapy, using the DNMT inhibitor decitabine (Vidaza) as an adjunct to NY-ESO-1 vaccine plus liposomal doxorubicin chemotherapy in 12 patients with relapsed EOC, with significant clinical benefit in the form of partial response disease or stabilization six of ten evaluable patients.

Similarly, the heterogeneous pattern of CTA (Cancer testis antigens) expression on tumor cells also occurs in acute myeloid leukemia and myelodysplasia and can be reduced, with corresponding therapeutic response increases, by DNA methyltransferase (DNMT) inhibition via azacitidine (AZA/Vidaza) and sodium valproate (VPA), with eight of the 11 patients with circulating MAGE (melanoma-associated antigens) CTLs (CD8+ cytotoxic T-lymphocyte) achieving a major clinical response after DNMT AZA/VPA inhibitor therapy [17]. This again confirms the use of epigenetic (DNMT and/or HDAC inhibition) therapy for the clinical modulation of intratumoral heterogeneity, in this case with clinical relevant improvement in T-cell therapy response, and comparable results in other malignancies (melanomas, RCC, etc.). [I have myself in the transformational oncology research I specialize in deployed natural dual DNMT/HDAC inhibitors to reduce the heterogeneity of triple negative breast cancer (TNBC) in patients in order to transform prognostically unfavorable disease into a prognostically favorable breast cancer subtype, namely hormone-positive (ER+) disease, with extensive live patient successes; this same approach has also been effective in overcoming resistance, in essence reverting the cancer epigenome of non-responsive cells to a drug-responsive state].

This is therefore a heterogeneity-modulation strategy for clinical response intent, and I will note that it suggests some exciting possibilities for other highly prognostically unfavorable malignancy subtypes: thus, I will point out the a recent Roswell Park Cancer Institute study has shown a comparable state of affairs in triple negative breast cancer [6,7,8,9,10,11], suggesting that a subset of patients with TNBC tumor expressive of NY-ESO-1 have particularly high inherent immunogenicity (that is, a measurably high spontaneous humoral immune response rate), making them an attractive population for cancer-germline/testis peptide vaccine trials, essentially NY-ESO-1-targeting heterogeneity-reductive studies. Thus we see here and in other confirmative studies [18,] that high levels of genetic heterogeneity are strongly associated with poor clinical outcomes. And note further that heterogeneity reduction also effects a corresponding reduction in therapeutic resistance, whether de novo or acquired.

Finally, the complexity of CSCs as instantiated by both heterogeneity and plasticity renders improbable that any one single agent will be efficient at targeting (as exemplified above), suggesting that an optimal strategy would be dual targeting of both CSCs and the non-stem populations in order to prevent the escape mechanism of reacquisition of stem-like characteristics by plasticity [29].

As to measurement of heterogeneity, fortunately for monitoring serial changes in tumor heterogeneity, there is mounting evidence that tracking circulating tumor cells (CTCs), via recent advances in the quantification and molecular characterization of CTCs with regard to important biomarkers, is both feasible and reasonable in investment as a tool for the measurement of the heterogeneity of the underlying tumors [24], and in addition the rapidly falling cost of even advanced technologies like next-generation sequencing (NGS) is making high-coverage DNA sequencing of clinically relevant cancer genes accessible at the point of care [25,26;21].

CLINICAL RELEVANCE OF THE HETEROGENEITY MODEL:

PREDICTIVE MARKERS

A second area of clinical relevance, besides therapeutic response and overcoming chemoresistance is in the area of predictive biomarkers. It is known that biomarker expression variability may reflect either genetic or nongenetic heterogeneity, and such biomarker heterogeneity may have clinicopathological and survival outcome relevance: for example, patients with tumors displaying heterogeneity in HER2 amplification are associated with a shorter DFS (disease-free survival) [16], strongly suggesting that measurements of heterogeneity for biomarkers provides clinically relevant information, in addition to the independent fact that genomic and phenotypic variability among tumor cells and the degree of intratumor heterogeneity may themselves be a prognostic factors [12,13,14,15]. The clinical lesson from the biomarker context is that heterogeneity can substantially confound and compromise the predictive power of biomarkers, and this entails as I see it that it will be absolutely necessary to incorporate measurements of intratumor heterogeneity during the validation of biomarkers in order to assure reliable predictive value and accuracy.

CLINICAL TRIALS OF HETEROGENEITY, AND THEIR CHALLENGES

REACT study [NCT01505400] will genomically evaluate all archived tumor samples from a cohort of molecularly profiled patients to assess heterogeneity and clonal evolution in several solid malignancies. And the interventional PROGENY (Prostate Cancer Genomic Heterogeneity) trial [NCT02022371] is designed to define the extent of inter/intra-tumor heterogeneity and its association with disease stage at diagnosis and Gleason grade, and to reconstruct inter/intra-tumor heterogeneity and clonal evolution occurring in men who have failed first- and second-line therapies (for localized and metastatic disease). In addition, the Breast Cancer Proteomics and Molecular Heterogeneity trial [NCT01840293] is examining the proteomic and molecular heterogeneity and associated characteristics of primary and recurrent/ metastatic breast tumors, while the BRAFV600E Intratumor Heterogeneity trial is examining comparable heterogeneity in thyroid cancer treated with TKIs, and the TRACERx trial [NCT01888601] is examining the dynamics of intratumor heterogeneity and associated phenotypics over time in NSCLC patients, tracking the relationship between intratumor heterogeneity and clinical outcome.

One problem for advancing the field of tumor and cellular heterogeneity further is that heterogeneity trials are highly protocol-challenging, especially if interventional (I sit on a panel exploring the optimal design of randomized trials for testing intratumoral and cellular heterogeneity), since the small population sizes typical and the frequent need to investigate the value of individualized heterogeneity-reductive therapy entails innovate designs. Here the N-of-1 clinical-trial design framework (Mod-N-of-1) is attractive. This design framework sequentially assess in the same patient the effects of different interventional agents, thus novelly using each individual patient as his or her own control (the comparison of treatment effect of the current matched intervention with that of the most recent prior intervention), in effect randomization is of the the order or scheduling in order to assess between and within person change and to investigate theoretically predicted mediators of those changes, as for instance being currently used in the WINTHER trial [NCT01856296] allowing for example to sequentially assess, in the same patient, the effects of different agents that may have antitumor activity against resistant clones. As has been argued convincingly this novel new-generation RCT may represent the ultimate strategy for individualizing/personalizing therapy [19,20,21] (I am a strong advocate), although I note that certain protocol designs are also adaptable in this context, such as the clustered RCT in which groups of subjects (as opposed to individual subjects) are randomized, that is, group-randomization, in that they can remove the problem of control group contamination which leads to biased estimates of effect size, as suggested by the new UK MRC Guidance on the Design and Development of Complex Trial Interventions [22,23], with choice of trial design in these complex individualized therapy interventions requiring special care and caution driven by the study goals and the characteristics and limitations of the interventions and the populations.

These and several other in-progress and planned trials should provide new insights into, and confirmation of, the operation and tenets of the new heterogeneity model.

It should therefore be clear from the above that these novel perspectives on multimodal heterogeneity have far-reaching clinical implications and will continue to occupy the frontiers of cancer research for the decade and beyond to come.

REFERENCES

1. Fidler IJ. 1978. Tumor heterogeneity and the biology of cancer invasion and metastasis. Cancer Res. 38:2651–60.

2. Heppner GH, Miller BE. 1983. Tumor heterogeneity: biological implications and therapeutic consequences. Cancer Metastasis Rev. 2:5–23.

3. Dexter DL, Kowalski HM, Blazar BA, Fligiel Z, Vogel R, Heppner GH. 1978. Heterogeneity of tumor cells from a single mouse mammary tumor. Cancer Res. 38:3174–81.

4. Odunsi K, Jungbluth AA, Stockert E, et al. NY-ESO-1 and LAGE-1 cancer-testis antigens are potential targets for immunotherapy in epithelial ovarian cancer. Cancer Res 2003 Sep 15; 63(18):6076-83.

5. Woloszynska-Read A, Mhawech-Fauceglia P, Yu J, Odunsi K, Karpf AR, et al. Intertumor and intratumor NY-ESO-1 expression heterogeneity is associated with promoter-specific and global DNA methylation status in ovarian cancer. Clin Cancer Res 2008 Jun 1; 14(11):3283-90.

6. Ademuyiwa FO, Bshara W, Attwood K, et al. NY-ESO-1 cancer testis antigen demonstrates high immunogenicity in triple negative breast cancer. PLoS One 2012; 7(6):e38783.

Stagg J, Allard B. Immunotherapeutic approaches in triple-negative breast cancer: latest research and clinical prospects. Ther Adv Med Oncol 2013; 5(3):169-81.

8. Karn T, Pusztai L, Ruckhäberle E, et al. Melanoma antigen family A identified by the bimodality index defines a subset of triple negative breast cancers as candidates for immune response augmentation. Eur J Cancer. 2012 Jan;48(1):12-23.

9. Cabezón T, Gromova I, Gromov P, et al. Proteomic profiling of triple-negative breast carcinomas in combination with a three-tier orthogonal technology approach identifies Mage-A4 as potential therapeutic target in estrogen receptor negative breast cancer. Mol Cell Proteomics 2013; 12(2):381-94.

10. Odunsi K, Matsuzaki J, James SR, et al. Epigenetic Potentiation of NY-ESO-1 Vaccine Therapy in Human Ovarian Cancer. Cancer Immunol Res January 2014 2; 37.

11. Polyak, K. Heterogeneity in breast cancer. J Clin Invest. 2011;121(10):3786–3788.

12. Barry WT, Kernagis DN, Dressman HK, Griffis RJ, Hunter JD, et al. Intratumor heterogeneity and precision of microarray-based predictors of breast cancer biology and clinical outcome. J. Clin. Oncol 2010; 28:2198–206.

13. YancovitzM, Litterman A, Yoon J, Ng E, Shapiro RL, et al. Intra- and inter-tumor heterogeneity of BRAF mutations in primary and metastatic melanoma. PLoS ONE 2012; 7:e29336.

14. NavinN,Krasnitz A, Rodgers L, Cook K, Meth J, et al. Inferring tumor progression from genomic heterogeneity. Genome Res 2010: 20:68–80.

15. Faratian D, Christiansen J,Gustavson M, JonesC, Scott C, et al. Heterogeneity mapping of protein expression in tumors using quantitative immunofluorescence. J. Vis. Exp 2011; 56:e3334.

16. Seol H, Lee HJ, Choi Y, Lee HE, Kim YJ, et al. 2012. Intratumoral heterogeneity of HER2 gene amplification in breast cancer: its clinicopathological significance. Mod. Pathol. 25:938–48.

17. Goodyear O, Agathanggelou A, Novitzky-Basso I, et al. Induction of a CD8+ T-cell response to the MAGE cancer testis antigen by combined treatment with azacitidine and sodium valproate in patients with acute myeloid leukemia and myelodysplasia. Blood 2010 Sep 16; 116(11):1908-18.

18. Maley CC, Galipeau PC, Finley JC,Wongsurawat VJ, Li X, et al. 2006. Genetic clonal diversity predicts progression to esophageal adenocarcinoma. Nat. Genet. 38:468–73.

19. Lillie EO, Patay B, Diamant J, Issell B, Topol EJ, Schork NJ. The n-of-1 clinical trial: the ultimate strategy for individualizing medicine? Per Med. 2011 Mar;8(2):161-173.

20. Gabler, N. B., Duan, N., Vohra, S. & Kravitz, R. L. N-of-1 trials in the medical literature: a systematic review. Med. Care 2011; 49, 761–768.

21. Bedard PL, Hansen AR, Ratain MJ, Siu LL. Tumour heterogeneity in the clinic. Nature 2013 Sep 19; 501(7467):355-64.

22. Craig P, Dieppe P, Macintyre S, … Medical Research Council Guidance. Developing and evaluating complex interventions: the new Medical Research Council guidance. BMJ 2008; 337:a1655.

23. Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M. MRC Population Health Sciences Research Network. Developing and evaluating complex interventions: the new Medical Research Council guidance. Int J Nurs Stud 2013; 50(5):587-92.

24. Hayes DF, Paoletti C. Circulating tumour cells: insights into tumour heterogeneity. J Intern Med 2013; 274(2):137-43.

25. Lipson, D. et al. Identification of new ALK and RET gene fusions from colorectal and lung cancer biopsies. Nature Med 2012; 18, 382–384.

26. Beltran, H. et al. Targeted next-generation sequencing of advanced prostate cancer identifies potential therapeutic targets and disease heterogeneity. Eur. Urol 2013; 63, 920–926.

27. Elshamy WM, Duhé RJ. Overview: cellular plasticity, cancer stem cells and metastasis. Cancer Lett 2013 Nov 28; 341(1):2-8.

28. Marjanovic ND, Weinberg RA, Chaffer CL. Cell plasticity and heterogeneity in cancer. Clin Chem 2013; 59(1):168-79.

29. R. French, R. Clarkson, The complex nature of breast cancer stem-like cells: Heterogeneity and plasticity. J Stem Cell Res Ther 2012, S7.

30. Wilting RH, Dannenberg JH. Epigenetic mechanisms in tumorigenesis, tumor cell heterogeneity and drug resistance. Drug Resist Updat 2012 Feb-Apr; 15(1-2):21-38.

Cancer stem cell (CSC) theory is highly accepted and even proved to describe the most of the traits related to or stemmed from tumor heterogeneity. However, there are some basic caveats in current understanding about the parameters that are needed for a cell to be called as CSC. For an instance, melanoma cells normally do not follow the currently accepted CSC model. You can see one of my recent article on this:

http://www.ncbi.nlm.nih.gov/pubmed/24615680

Here you find something interesting......

Article Tumor heterogeneity and cancer stem cell paradigm: Updates i...

Tumour heterogeneity describes a state in which there are significant variations in a particular cancer cell population. It's often displayed not just in gene expression but also in cancer proliferative as well as metastatic tendencies. We can look at the topic from several angles. First is the fact that any condition that brings about instability in the cancer genome is a potential trigger for heterogeneity. Second is the role of some non-genetic or epigenetic influences on the cancer cell populations. Use of chemotherapeutic agents for instance has been identified as a potential source of selection pressure on the tumour cell population, allowing those cells with inherent resistance potential to continue to grow and proliferate. This is one possible mechanism behind the evolution of resistant cancer sub-clones. Whereas Cancer Stem Cell Model is very popular in literature, Clonal Evolution is equally very relevant in tumour heterogeneity.

Shaobo:

This is an area of my own research specialization, but given the fine contributions above, I will restrict my attention to two primary themes that should add some further nuance and precision to the discussion: (1) the radical divergence of the new model of heterogeneity from its predecessor; (2) and as always, the clinical relevance of the new model, in particular in the domains of therapeutic response, anti-resistance, and predictive biomarkers, with examples drawn from human clinical data.

The recognition of intratumor heterogeneity for cellular phenotypes is indeed ancient history. In fact it began over 150+years ago: the great Ur-pathologist Rudolf Virchow observed intratumoral pleomorphism of cancer cells, namely morphological heterogeneity of contained malignant cells within individual tumors, in the mid-1800's (circa 1859), followed by demonstrations of the intratumoral functional and genetic heterogeneity, especially the demonstrations over 30 to 40 years ago of distinct subpopulations of cancer cells within tumors, differentiated as to tumorigenicity, therapeutic resistance, and metastatic potential [1,2], including demonstrations by molecular analysis of specifically genetic variability among individual cancer cells [3].

THE NEW HETEROGENEITY MODEL:

The differences however from these historical roots are vast, and new focus however has shifted dramatically and uniquely to:

(1) CELLULAR HETEROGENEITY FOCUS:

Intratumor phenotypic heterogeneity at the single-cell level, aka cancer cell heterogeneity (or cell-intrinsic heterogeneity).

(2) MULTIMODAL HETEROGENEITY:

This is now within the domains of not just tumor biological and histopathological domains, but also in domains of:

- GENETIC HETEROGENEITY (including heterogeneity of normal stem/progenitor cells),

- EPIGENETIC HETEROGENEITY given that epigenetics has been recognized as an important factor in generating non-genetic heterogeneity, an area that has received to date insufficient attention from heterogeneity theorists but stands to be one of the most clinically relevant domains of application [30;13]. From the epigenetic perspective on heterogeneity, combination anticancer therapeutics of cytotoxic or biological agents with epigenetic agents, in particular HDAC and DNMT inhibitors, enables an multiple-angle assault on heterogeneous tumor cell populations that minimizes the evasion opportunities for drug-resistant cells to escape and generate a drug-resistant tumor.

- DETERMINISTIC CELLULAR HETEROGENEITY (multiple fairly stable phenotypic states)

- STOCHASTIC CELLULAR HETEROGENEITY (transient differences in phenotypes between isogenic cells that share the same deterministic phenotypic state),

- CANCER STEM CELL (CSC) HETEROGENEITY, and

- METASTATIC CELLULAR HETEROGENEITY,

- TUMOR MICROENVIRONMENT HETEROGENEITY: there is a new recognition of tumor microenvironment heterogeneity, and with not only intratumor heterogeneity of tumor cells, but also of stromal cells and non-cellular components of microenvironments, and note that the heterogeneity of tumor microenvironments translates into heterogeneous selective pressures experienced by the tumor cells themselves.

(3) CONTINUUM MODEL:

A new "continuum model" of intratumoral heterogeneity in which resident cells reside in different states of stemness or differentiation, on a spectrum of degrees.

(4) CELLULAR PLASTICITY:

This too is a new focus and direction, and represents a wholesale rejection of the classical perspective that cancer stem cell (CSC)-to–non-CSC conversion is a strictly unidirectional process: the plasticity postulate within the cancer cell heterogeneity viewpoint represents as new model of bidirectional interconversions between non-CSCs and CSCs, entailing that non-CSCs can continually create CSC populations throughout tumorigenesis. In this sense, the bold new hypothesis is that the stem cell state is reentrant, with molecular and cellular mechanisms enabling the reentering of the stemness state [27,28], and newly viewing cellular differentiation not as, traditionally, a unidirectional process, but a plastic process where cancer cells can dedifferentiate into more primitive, stem-like phenotypes, these plastic phenotypic shifts finally helping to account the discontinuous behavior of cancer evidenced in some cancers remaining dormant for extended durations (months or years) after therapy, only to relapse later.

THE TUMOR ECOSYSTEM

This gives rise to a model of heterogeneous tumors as complex ecosystems, wherein even a minor tumor subpopulation can influence global tumor growth, that is growth of the entire tumor, actively maintaining tumor heterogeneity and in turn possibly confounding predictive biomarkers and adversely facilitating therapeutic resistance.

CLINICAL RELEVANCE OF THE HETEROGENEITY MODEL:

THERAPEUTIC RESPONSE

One area of critical therapeutic relevance arises from a strategy of heterogeneity modulation, the targeted reduction in tumoral and cellular heterogeneity via heterogeneity-reductive pathways, which is one species of what I have otherwise termed transformational oncology. So as just one example (there are dozens), I will take NY-ESO-1:

NY-ESO-1 is an immunogenic cancer-germline/testis antigen aberrantly expressed in several human malignancies, including epithelial ovarian cancer (EOC) where it is deployed for immunotherapy, and it's known that a frank majority of NY-ESO-1–positive EOC tumors display a heterogeneous expression pattern of this antigen, causing immune and clinical response to NY-ESO-1 peptide vaccine therapy be limited secondary to this heterogeneous expression pattern among (inter-) and within (intra-) tumors [4,5].

Where the clinical therapeutic connection arise is that intratumoral NY-ESO-1 expression is dependent on promoter methylation (both promoter-specific and global DNA methylation status). And as the seminal Roswell Park Cancer Institute study [5] demonstrated, DNA hypomethylation induction with (5-)azacitidine (Vidaza), a DNA methylation inhibitor that is a special form of DNMT (DNA methyltransferase) inhibitor, can restore NY-ESO-1 expression in nonresponder cells. In essence, what we have here is that promoter methylation regulates NY-ESO-1 expression heterogeneity in epithelial ovarian cancer (EOC), and that either treatment with a DNMT inhibitor (azacitidine ) or an HDAC inhibitor (decitabine/Dacogen), functionally restores NY-ESO-1 expression in nonexpressing (nonresponder) EOC cell lines. And I am glad to report that this has recently [this January, 2014] received clinical confirmation in a Phase I trial [10] showing that DNA methyltransferase (DNMT) inhibitors augment NY-ESO-1 vaccine therapy, using the DNMT inhibitor decitabine (Vidaza) as an adjunct to NY-ESO-1 vaccine plus liposomal doxorubicin chemotherapy in 12 patients with relapsed EOC, with significant clinical benefit in the form of partial response disease or stabilization six of ten evaluable patients.

Similarly, the heterogeneous pattern of CTA (Cancer testis antigens) expression on tumor cells also occurs in acute myeloid leukemia and myelodysplasia and can be reduced, with corresponding therapeutic response increases, by DNA methyltransferase (DNMT) inhibition via azacitidine (AZA/Vidaza) and sodium valproate (VPA), with eight of the 11 patients with circulating MAGE (melanoma-associated antigens) CTLs (CD8+ cytotoxic T-lymphocyte) achieving a major clinical response after DNMT AZA/VPA inhibitor therapy [17]. This again confirms the use of epigenetic (DNMT and/or HDAC inhibition) therapy for the clinical modulation of intratumoral heterogeneity, in this case with clinical relevant improvement in T-cell therapy response, and comparable results in other malignancies (melanomas, RCC, etc.). [I have myself in the transformational oncology research I specialize in deployed natural dual DNMT/HDAC inhibitors to reduce the heterogeneity of triple negative breast cancer (TNBC) in patients in order to transform prognostically unfavorable disease into a prognostically favorable breast cancer subtype, namely hormone-positive (ER+) disease, with extensive live patient successes; this same approach has also been effective in overcoming resistance, in essence reverting the cancer epigenome of non-responsive cells to a drug-responsive state].

This is therefore a heterogeneity-modulation strategy for clinical response intent, and I will note that it suggests some exciting possibilities for other highly prognostically unfavorable malignancy subtypes: thus, I will point out the a recent Roswell Park Cancer Institute study has shown a comparable state of affairs in triple negative breast cancer [6,7,8,9,10,11], suggesting that a subset of patients with TNBC tumor expressive of NY-ESO-1 have particularly high inherent immunogenicity (that is, a measurably high spontaneous humoral immune response rate), making them an attractive population for cancer-germline/testis peptide vaccine trials, essentially NY-ESO-1-targeting heterogeneity-reductive studies. Thus we see here and in other confirmative studies [18,] that high levels of genetic heterogeneity are strongly associated with poor clinical outcomes. And note further that heterogeneity reduction also effects a corresponding reduction in therapeutic resistance, whether de novo or acquired.

Finally, the complexity of CSCs as instantiated by both heterogeneity and plasticity renders improbable that any one single agent will be efficient at targeting (as exemplified above), suggesting that an optimal strategy would be dual targeting of both CSCs and the non-stem populations in order to prevent the escape mechanism of reacquisition of stem-like characteristics by plasticity [29].

As to measurement of heterogeneity, fortunately for monitoring serial changes in tumor heterogeneity, there is mounting evidence that tracking circulating tumor cells (CTCs), via recent advances in the quantification and molecular characterization of CTCs with regard to important biomarkers, is both feasible and reasonable in investment as a tool for the measurement of the heterogeneity of the underlying tumors [24], and in addition the rapidly falling cost of even advanced technologies like next-generation sequencing (NGS) is making high-coverage DNA sequencing of clinically relevant cancer genes accessible at the point of care [25,26;21].

CLINICAL RELEVANCE OF THE HETEROGENEITY MODEL:

PREDICTIVE MARKERS

A second area of clinical relevance, besides therapeutic response and overcoming chemoresistance is in the area of predictive biomarkers. It is known that biomarker expression variability may reflect either genetic or nongenetic heterogeneity, and such biomarker heterogeneity may have clinicopathological and survival outcome relevance: for example, patients with tumors displaying heterogeneity in HER2 amplification are associated with a shorter DFS (disease-free survival) [16], strongly suggesting that measurements of heterogeneity for biomarkers provides clinically relevant information, in addition to the independent fact that genomic and phenotypic variability among tumor cells and the degree of intratumor heterogeneity may themselves be a prognostic factors [12,13,14,15]. The clinical lesson from the biomarker context is that heterogeneity can substantially confound and compromise the predictive power of biomarkers, and this entails as I see it that it will be absolutely necessary to incorporate measurements of intratumor heterogeneity during the validation of biomarkers in order to assure reliable predictive value and accuracy.

CLINICAL TRIALS OF HETEROGENEITY, AND THEIR CHALLENGES

REACT study [NCT01505400] will genomically evaluate all archived tumor samples from a cohort of molecularly profiled patients to assess heterogeneity and clonal evolution in several solid malignancies. And the interventional PROGENY (Prostate Cancer Genomic Heterogeneity) trial [NCT02022371] is designed to define the extent of inter/intra-tumor heterogeneity and its association with disease stage at diagnosis and Gleason grade, and to reconstruct inter/intra-tumor heterogeneity and clonal evolution occurring in men who have failed first- and second-line therapies (for localized and metastatic disease). In addition, the Breast Cancer Proteomics and Molecular Heterogeneity trial [NCT01840293] is examining the proteomic and molecular heterogeneity and associated characteristics of primary and recurrent/ metastatic breast tumors, while the BRAFV600E Intratumor Heterogeneity trial is examining comparable heterogeneity in thyroid cancer treated with TKIs, and the TRACERx trial [NCT01888601] is examining the dynamics of intratumor heterogeneity and associated phenotypics over time in NSCLC patients, tracking the relationship between intratumor heterogeneity and clinical outcome.

One problem for advancing the field of tumor and cellular heterogeneity further is that heterogeneity trials are highly protocol-challenging, especially if interventional (I sit on a panel exploring the optimal design of randomized trials for testing intratumoral and cellular heterogeneity), since the small population sizes typical and the frequent need to investigate the value of individualized heterogeneity-reductive therapy entails innovate designs. Here the N-of-1 clinical-trial design framework (Mod-N-of-1) is attractive. This design framework sequentially assess in the same patient the effects of different interventional agents, thus novelly using each individual patient as his or her own control (the comparison of treatment effect of the current matched intervention with that of the most recent prior intervention), in effect randomization is of the the order or scheduling in order to assess between and within person change and to investigate theoretically predicted mediators of those changes, as for instance being currently used in the WINTHER trial [NCT01856296] allowing for example to sequentially assess, in the same patient, the effects of different agents that may have antitumor activity against resistant clones. As has been argued convincingly this novel new-generation RCT may represent the ultimate strategy for individualizing/personalizing therapy [19,20,21] (I am a strong advocate), although I note that certain protocol designs are also adaptable in this context, such as the clustered RCT in which groups of subjects (as opposed to individual subjects) are randomized, that is, group-randomization, in that they can remove the problem of control group contamination which leads to biased estimates of effect size, as suggested by the new UK MRC Guidance on the Design and Development of Complex Trial Interventions [22,23], with choice of trial design in these complex individualized therapy interventions requiring special care and caution driven by the study goals and the characteristics and limitations of the interventions and the populations.

These and several other in-progress and planned trials should provide new insights into, and confirmation of, the operation and tenets of the new heterogeneity model.

It should therefore be clear from the above that these novel perspectives on multimodal heterogeneity have far-reaching clinical implications and will continue to occupy the frontiers of cancer research for the decade and beyond to come.

REFERENCES

1. Fidler IJ. 1978. Tumor heterogeneity and the biology of cancer invasion and metastasis. Cancer Res. 38:2651–60.

2. Heppner GH, Miller BE. 1983. Tumor heterogeneity: biological implications and therapeutic consequences. Cancer Metastasis Rev. 2:5–23.

3. Dexter DL, Kowalski HM, Blazar BA, Fligiel Z, Vogel R, Heppner GH. 1978. Heterogeneity of tumor cells from a single mouse mammary tumor. Cancer Res. 38:3174–81.

4. Odunsi K, Jungbluth AA, Stockert E, et al. NY-ESO-1 and LAGE-1 cancer-testis antigens are potential targets for immunotherapy in epithelial ovarian cancer. Cancer Res 2003 Sep 15; 63(18):6076-83.

5. Woloszynska-Read A, Mhawech-Fauceglia P, Yu J, Odunsi K, Karpf AR, et al. Intertumor and intratumor NY-ESO-1 expression heterogeneity is associated with promoter-specific and global DNA methylation status in ovarian cancer. Clin Cancer Res 2008 Jun 1; 14(11):3283-90.

6. Ademuyiwa FO, Bshara W, Attwood K, et al. NY-ESO-1 cancer testis antigen demonstrates high immunogenicity in triple negative breast cancer. PLoS One 2012; 7(6):e38783.

Stagg J, Allard B. Immunotherapeutic approaches in triple-negative breast cancer: latest research and clinical prospects. Ther Adv Med Oncol 2013; 5(3):169-81.

8. Karn T, Pusztai L, Ruckhäberle E, et al. Melanoma antigen family A identified by the bimodality index defines a subset of triple negative breast cancers as candidates for immune response augmentation. Eur J Cancer. 2012 Jan;48(1):12-23.

9. Cabezón T, Gromova I, Gromov P, et al. Proteomic profiling of triple-negative breast carcinomas in combination with a three-tier orthogonal technology approach identifies Mage-A4 as potential therapeutic target in estrogen receptor negative breast cancer. Mol Cell Proteomics 2013; 12(2):381-94.

10. Odunsi K, Matsuzaki J, James SR, et al. Epigenetic Potentiation of NY-ESO-1 Vaccine Therapy in Human Ovarian Cancer. Cancer Immunol Res January 2014 2; 37.

11. Polyak, K. Heterogeneity in breast cancer. J Clin Invest. 2011;121(10):3786–3788.

12. Barry WT, Kernagis DN, Dressman HK, Griffis RJ, Hunter JD, et al. Intratumor heterogeneity and precision of microarray-based predictors of breast cancer biology and clinical outcome. J. Clin. Oncol 2010; 28:2198–206.

13. YancovitzM, Litterman A, Yoon J, Ng E, Shapiro RL, et al. Intra- and inter-tumor heterogeneity of BRAF mutations in primary and metastatic melanoma. PLoS ONE 2012; 7:e29336.

14. NavinN,Krasnitz A, Rodgers L, Cook K, Meth J, et al. Inferring tumor progression from genomic heterogeneity. Genome Res 2010: 20:68–80.

15. Faratian D, Christiansen J,Gustavson M, JonesC, Scott C, et al. Heterogeneity mapping of protein expression in tumors using quantitative immunofluorescence. J. Vis. Exp 2011; 56:e3334.

16. Seol H, Lee HJ, Choi Y, Lee HE, Kim YJ, et al. 2012. Intratumoral heterogeneity of HER2 gene amplification in breast cancer: its clinicopathological significance. Mod. Pathol. 25:938–48.

17. Goodyear O, Agathanggelou A, Novitzky-Basso I, et al. Induction of a CD8+ T-cell response to the MAGE cancer testis antigen by combined treatment with azacitidine and sodium valproate in patients with acute myeloid leukemia and myelodysplasia. Blood 2010 Sep 16; 116(11):1908-18.

18. Maley CC, Galipeau PC, Finley JC,Wongsurawat VJ, Li X, et al. 2006. Genetic clonal diversity predicts progression to esophageal adenocarcinoma. Nat. Genet. 38:468–73.

19. Lillie EO, Patay B, Diamant J, Issell B, Topol EJ, Schork NJ. The n-of-1 clinical trial: the ultimate strategy for individualizing medicine? Per Med. 2011 Mar;8(2):161-173.

20. Gabler, N. B., Duan, N., Vohra, S. & Kravitz, R. L. N-of-1 trials in the medical literature: a systematic review. Med. Care 2011; 49, 761–768.

21. Bedard PL, Hansen AR, Ratain MJ, Siu LL. Tumour heterogeneity in the clinic. Nature 2013 Sep 19; 501(7467):355-64.

22. Craig P, Dieppe P, Macintyre S, … Medical Research Council Guidance. Developing and evaluating complex interventions: the new Medical Research Council guidance. BMJ 2008; 337:a1655.

23. Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M. MRC Population Health Sciences Research Network. Developing and evaluating complex interventions: the new Medical Research Council guidance. Int J Nurs Stud 2013; 50(5):587-92.

24. Hayes DF, Paoletti C. Circulating tumour cells: insights into tumour heterogeneity. J Intern Med 2013; 274(2):137-43.

25. Lipson, D. et al. Identification of new ALK and RET gene fusions from colorectal and lung cancer biopsies. Nature Med 2012; 18, 382–384.

26. Beltran, H. et al. Targeted next-generation sequencing of advanced prostate cancer identifies potential therapeutic targets and disease heterogeneity. Eur. Urol 2013; 63, 920–926.

27. Elshamy WM, Duhé RJ. Overview: cellular plasticity, cancer stem cells and metastasis. Cancer Lett 2013 Nov 28; 341(1):2-8.

28. Marjanovic ND, Weinberg RA, Chaffer CL. Cell plasticity and heterogeneity in cancer. Clin Chem 2013; 59(1):168-79.

29. R. French, R. Clarkson, The complex nature of breast cancer stem-like cells: Heterogeneity and plasticity. J Stem Cell Res Ther 2012, S7.

30. Wilting RH, Dannenberg JH. Epigenetic mechanisms in tumorigenesis, tumor cell heterogeneity and drug resistance. Drug Resist Updat 2012 Feb-Apr; 15(1-2):21-38.

Dear doctor,

Your comments are greatly appreciated. Categories of heterogeneity we are encountering everyday were clearly listed. The heterogeneity is important because it is clinical relevant and critical for the patient management. It partially explains the selection theory and field carcinogenesis. Heterogeneity is also critical for the targeted therapy and even genetic diagnosis.

Appreciated for the answers that look insight of many aspect of the tumor heterogeneity. The question is clinical relevant and related to the route of carcinogenesis. From my work on bladder and prostate cancer, it seemed likely that at least these cancers have multiple clonal origin and the metastatic lesions mostly associated to a single clone. That indicated that the properties of those clones are biologically different. From this point of view it is more appreciated to the theory of field carcinogenesis/cancer stem cells.

Dear Shaobo,

Here is my response just a little bit shorter than Constandine's ones ...

Theories remain thories and indeed som good writers like to write textbooks about various theories.

From my very humble point of view, some up-to-date top-ranked articles are far better to "forge" my opinion than long stories.

Here attached are some articles that I think could contribute to your understanding about the question you raise.

Best regards

Robert