Framework models of tumor dormancy from patient-derived observations

Unusually long latency periods between the treatment of primary tumors and metastatic recurrences are commonly thought to prove the existence and relevance of clinical tumor dormancy. However, careful consideration of disease courses and cancer growth rates leads to the conclusion that clinical dormancy may be everything from non-existent to much more frequent than originally thought. On the other hand, cellular dormancy defined as a non-productive state of disseminated tumor cells is very frequent, while homeostatic mechanisms such as angiogenic and immunological control contribute to the chronicity of cancer. This review attempts to provide a conceptual framework for the study of dormancy, which may guide clinical and experimental research.     

The Problem of Cancer Dormancy: Understanding the Basic Mechanisms and Identifying Therapeutic Opportunities

The hiatus observed in the progression of cancer after diagnosis and treatment in a large proportion of patients has led to the notion that a state of cancer dormancy must exist during tumor progression. However, research on this stage of cancer has been limited due to the lack of appropriate models and clinical correlates. Fortunately, the last decade has seen the development of new cancer dormancy models, whole animal and intravital imaging techniques and the molecular characterization of minimal residual disease. These studies enabled researchers to reveal intriguing mechanisms and molecular determinants that define tumor dormancy. It is imperative to understand the basic mechanisms of dormancy, as this will accelerate the development of new markers of progression and novel therapeutic opportunities to induce dormancy and/or eradicate dormant disease. This issue of Cell Cycle includes a “Spotlight on Cancer Dormancy” highlighting major contributions to the field of cancer dormancy from basic and clinical studies. We anticipate that this will initiate a forum of discussion on the problem of cancer dormancy and stimulate investigators to study this rather unexplored but undeniably relevant clinical stage of cancer progression.     

Dormancy signatures and metastasis in estrogen receptor positive and negative breast cancer

Breast cancers can recur after removal of the primary tumor and treatment to eliminate remaining tumor cells. Recurrence may occur after long periods of time during which there are no clinical symptoms. Tumor cell dormancy may explain these prolonged periods of asymptomatic residual disease and treatment resistance. We generated a dormancy gene signature from published experimental models and applied it to both breast cancer cell line expression data as well as four published clinical studies of primary breast cancers. We found that estrogen receptor (ER) positive breast cell lines and primary tumors have significantly higher dormancy signature scores (P<0.0000001) than ER- cell lines and tumors. In addition, a stratified analysis combining all ER+ tumors in four studies indicated 2.1 times higher hazard of recurrence among patients whose tumors had low dormancy scores (LDS) compared to those whose tumors had high dormancy scores (HDS) (p<0.000005). The trend was shown in all four individual studies. Suppression of two dormancy genes, BHLHE41 and NR2F1, resulted in increased in vivo growth of ER positive MCF7 cells. The patient data analysis suggests that disseminated ER positive tumor cells carrying a dormancy signature are more likely to undergo prolonged dormancy before resuming metastatic growth. Furthermore, genes identified with this approach might provide insight into the mechanisms of dormancy onset and maintenance as well as dormancy models using human breast cancer cell lines.     

Systemic Cancer Progression and Tumor Dormancy: Mathematical Models Meet Single Cell Genomics

Metastatic progression is thought to result from genetically advanced ?fully-malignant“ tumor cells. Within the concept the prevailing view holds that such cells disseminate mostly from large tumors and are capable of growing into metastases once they arrive at a distant site. Support for this scenario comes from numerous mouse models in which transplanted tumor cells grow into metastases within days or weeks. However, the assumption of such fully-malignant disseminating cells in human cancer is misleading and is neither supported by mathematical modeling of survival data from cancer patients nor by ex-vivo genomic data from disseminated cancer cells. For example, in breast cancer the growth of metastases is highly homogeneous and takes on average six years, the number of disseminated tumor cells before diagnosis of metastasis is similar for different tumor stages, and the genomic aberrations of disseminated cancer cells do rarely correspond to those in the primary tumor. Since these facts question conventional concepts of metastatic progression we provide a model of cancer progression in which time considerations and direct ex-vivo data form a starting point. In the proposed model tumor dormancy is a characteristic of almost all migrated tumor cells and metastatic growth is a rare, stochastic, evolutionary process of selection and mutation of cells that often disseminate shortly after transformation at the primary site.     

Role of Angiogenesis in Human Tumor Dormancy: Animal Models of the Angiogenic Switch

Tumor progression depends on sequential events, including a switch to the angiogenic phenotype (i.e. initial recruitment of blood vessels). Failure of a microscopic tumor to complete one or more early steps in this process may lead to delayed clinical manifestation of the cancer. Microscopic human cancers can remain in an asymptomatic, non-detectable, and occult state for the life of a person. Clinical and experimental evidence suggest that human tumors can persist for long periods of time as microscopic lesions that are in a state of dormancy (i.e. not expanding in tumor mass). Because it is well established that tumor growth beyond the size of 1-2 mm is angiogenesis-dependent, we hypothesized that presentation of large tumors is attributed to a switch to the angiogenic phenotype in otherwise microscopic, dormant tumors. Although clinically important, the biology of human tumor dormancy is poorly understood. The development of animal models which recapitulate the clinically observed timing and proportion of dormant tumors which switch to the angiogenic phenotype are reviewed here. The contributing molecular mechanisms involved in the angiogenic switch and different strategies for isolation of both angiogenic and nonangiogenic tumor cell populations from otherwise heterogeneous human tumor cell lines or surgical specimens are also summarized. Several imaging techniques have been utilized for the qualitative and quantitative detection of microscopic tumors in mice and their strengths and limitations are discussed. The animal models employed here permitted further studies of the angiogenic switch. These models also allowed development of an angiogenesis-based panel of blood and urine biomarkers that can be quantified and used to detect microscopic tumors before or during the angiogenic switch. If the information obtained from these animal models is translatable to the clinic, it may be possible in the future to liberate the management of cancer from a dependency on anatomical site years before it becomes symptomatic and detectable.     

Angiosuppression and chemotherapy: strategies aimed at their integration in cancer patients

A number of antiangiogenic agents have been developed as pharmaceuticals and are currently being tested in clinical studies. Potential strategies to enhance the activity of angiogenesis inhibitors could be to combine them, or better still, to administer them either sequentially or concurrently with cytotoxic drugs. Chemotherapy would be a more appropriate initial choice for patients with advanced disease since cytostatic agents can induce a fast regression of the tumor and cancer-related symptoms. Antiangiogenic treatment could be used after chemotherapy in patients who achieve disease remission to prolong the time to progression, the symptom-free interval and the overall survival. Antiangiogenic treatment is likely to attain an important role in the adjuvant setting. In fact, it could be used for prolonged periods after radical surgery to maintain dormancy of residual tumor cells. In spite of these promising preclinical data, several points need to be clarified before the initiation of clinical trials. In fact, certain misconceptions may interfere with their optimum design and result analysis.     

Bayesian modeling of incidence and progression of disease from cross-sectional data

In the absence of longitudinal data, the current presence and severity of disease can be measured for a sample of individuals to investigate factors related to disease incidence and progression. In this article, Bayesian discrete-time stochastic models are developed for inference from cross-sectional data consisting of the age at first diagnosis, the current presence of disease, and one or more surrogates of disease severity. Semiparametric models are used for the age-specific hazards of onset and diagnosis, and a normal underlying variable approach is proposed for modeling of changes with latency time in disease severity. The model accommodates multiple surrogates of disease severity having different measurement scales and heterogeneity among individuals in disease progression. A Markov chain Monte Carlo algorithm is described for posterior computation, and the methods are applied to data from a study of uterine leiomyoma.     

Cancer progression by non-clonal chromosome aberrations

The establishment of the correct conceptual framework is vital to any scientific discipline including cancer research. Influenced by hematologic cancer studies, the current cancer concept focuses on the stepwise patterns of progression as defined by specific recurrent genetic aberrations. This concept has faced a tough challenge as the majority of cancer cases follow non-linear patterns and display stochastic progression. In light of the recent discovery that genomic instability is directly linked to stochastic non-clonal chromosome aberrations (NCCAs), and that cancer progression can be characterized as a dynamic relationship between NCCAs and recurrent clonal chromosome aberrations (CCAs), we propose that the dynamics of NCCAs is a key element for karyotypic evolution in solid tumors. To support this viewpoint, we briefly discuss various basic elements responsible for cancer initiation and progression within an evolutionary context. We argue that even though stochastic changes can be detected at various levels of genetic organization, such as at the gene level and epigenetic level, it is primarily detected at the chromosomal or genome level. Thus, NCCA-mediated genomic variation plays a dominant role in cancer progression. To further illustrate the involvement of NCCA/CCA cycles in the pattern of cancer evolution, four cancer evolutionary models have been proposed based on the comparative analysis of karyotype patterns of various types of cancer.     

A Cellular Automaton Model for Tumor Dormancy: Emergence of a Proliferative Switch

Malignant cancers that lead to fatal outcomes for patients may remain dormant for very long periods of time. Although individual mechanisms such as cellular dormancy, angiogenic dormancy and immunosurveillance have been proposed, a comprehensive understanding of cancer dormancy and the “switch” from a dormant to a proliferative state still needs to be strengthened from both a basic and clinical point of view. Computational modeling enables one to explore a variety of scenarios for possible but realistic microscopic dormancy mechanisms and their predicted outcomes. The aim of this paper is to devise such a predictive computational model of dormancy with an emergent “switch” behavior. Specifically, we generalize a previous cellular automaton (CA) model for proliferative growth of solid tumor that now incorporates a variety of cell-level tumor-host interactions and different mechanisms for tumor dormancy, for example the effects of the immune system. Our new CA rules induce a natural “competition” between the tumor and tumor suppression factors in the microenvironment. This competition either results in a “stalemate” for a period of time in which the tumor either eventually wins (spontaneously emerges) or is eradicated; or it leads to a situation in which the tumor is eradicated before such a “stalemate” could ever develop. We also predict that if the number of actively dividing cells within the proliferative rim of the tumor reaches a critical, yet low level, the dormant tumor has a high probability to resume rapid growth. Our findings may shed light on the fundamental understanding of cancer dormancy.     

A new view of radiation-induced cancer: integrating short- and long-term processes. Part I: Approach

Mathematical models of radiation carcinogenesis are important for understanding mechanisms and for interpreting or extrapolating risk. There are two classes of such models: (1) long-term formalisms that track pre-malignant cell numbers throughout an entire lifetime but treat initial radiation dose–response simplistically and (2) short-term formalisms that provide a detailed initial dose–response even for complicated radiation protocols, but address its modulation during the subsequent cancer latency period only indirectly. We argue that integrating short- and long-term models is needed. As an example of this novel approach, we integrate a stochastic short-term initiation/inactivation/repopulation model with a deterministic two-stage long-term model. Within this new formalism, the following assumptions are implemented: radiation initiates, promotes, or kills pre-malignant cells; a pre-malignant cell generates a clone, which, if it survives, quickly reaches a size limitation; the clone subsequently grows more slowly and can eventually generate a malignant cell; the carcinogenic potential of pre-malignant cells decreases with age.     

Joint modelling of repeated transitions in follow-up data--a case study on breast cancer data

In longitudinal studies where time to a final event is the ultimate outcome often information is available about intermediate events the individuals may experience during the observation period. Even though many extensions of the Cox proportional hazards model have been proposed to model such multivariate time-to-event data these approaches are still very rarely applied to real datasets. The aim of this paper is to illustrate the application of extended Cox models for multiple time-to-event data and to show their implementation in popular statistical software packages. We demonstrate a systematic way of jointly modelling similar or repeated transitions in follow-up data by analysing an event-history dataset consisting of 270 breast cancer patients, that were followed-up for different clinical events during treatment in metastatic disease. First, we show how this methodology can also be applied to non Markovian stochastic processes by representing these processes as "conditional" Markov processes. Secondly, we compare the application of different Cox-related approaches to the breast cancer data by varying their key model components (i.e. analysis time scale, risk set and baseline hazard function). Our study showed that extended Cox models are a powerful tool for analysing complex event history datasets since the approach can address many dynamic data features such as multiple time scales, dynamic risk sets, time-varying covariates, transition by covariate interactions, autoregressive dependence or intra-subject correlation.     

Use of transgenic mice as models for prostate cancer chemoprevention

Prostate cancer is a long latency type of tumor that usually develops in men older than 50 years of age. Prostate epithelial neoplasia (PIN), the initial malignant lesion, progresses to invasive carcinoma over the course of years. Because of the particular features of prostate carcinogenesis, this type of tumor may represent a paradigm for cancer prevention. Several clinical trials have evaluated the effect of different compounds on prostate tumor development, including finasteride, selenium, vitamin E, and carotenes. Although some results are promising, no conclusive data have been achieved as to recommend any of these compounds as preventive agents. Results from some trials, such as SELECT, where supplementation of selenium and/or vitamin-E was used, have been rather disappointing. However, many novel chemopreventive agents that target different cancer-related pathways are being developed lately. Appropriate animal models (in particular, genetically modified mice) are being used to assess the efficacy of these novel compounds. The transgenic adenocarcinoma of the mouse prostate (TRAMP) model has been validated as an accurate model to test a variety of preventive agents. Genistein, alpha-difluoromethylornithine, toremifene, R-flurbiprofen, celecoxib, and green tea polyphenols have been shown to prevent prostate cancer development in TRAMP mice. In conclusion, new chemopreventive compounds which are effective in animal models are likely to be tested soon in clinical trials, with the final goal of reducing prostate cancer incidence in men.     

Tumour necrosis factor and cancer

TNF is a cytokine whose diverse actions are dependent on the local microenvironment. As a member of the cytokine network, TNF plays an important role in infection and inflammation, but excessive and deregulated production can contribute to disease processes. Likewise in malignant disease, TNF may have a role in cancer therapy and contribute to host response against tumours, but it may also be involved in the progression and spread of the cancer. In experimental models, recombinant TNF can induce significant haemorrhagic necrosis, localised to the tumour vasculature and specific tumour immunity. Although the historical background and preclinical data are promising, systemic therapy with TNF in human cancer has proved highly toxic and is inactive against all tumour types so far tested. Local therapy, particularly isolated limb perfusion, has resulted in complete and long lasting tumour regressions with necrotic activity confined solely to the tumour vascular bed. However, in several animal models, TNF contributes to malignant progression and there is evidence that TNF may have autocrine or paracrine actions in human ovarian cancer.     

The tumor macroenvironment and systemic regulation of breast cancer progression

Breast cancer is the most common malignancy among women worldwide and is the most common cause of death for women between 35 and 50 years of age. Women with breast cancer are at risk of developing metastases for their entire lifetime and, despite local and systemic therapies, approximately 30% of breast cancer patients will relapse (Jemal et al., 2010). Nearly all breast cancer related deaths are due to metastatic disease, even though metastasis is considered to be an inefficient process. In some cases, tumor cells disseminate from primary sites at an early stage, but remain indolent for protracted periods of time before becoming overt, life-threatening tumors. Little is known about the mechanisms that cause these indolent tumors to grow into malignant disease. Because of this gap in our understanding, we are unable to predict which breast cancer patients are likely to experience disease relapse or develop metastases years after treatment of their primary tumor. A better understanding of the mechanisms and signals involved in the exit of tumor cells from dormancy would not only allow for more accurate selection of patients that would benefit from systemic therapy, but could also lead to the development of more targeted therapies to inhibit the signals that promote disease progression. In this review, we address the systemic, or "macroenvironmental", contribution to tumor initiation and progression and what is known about how a pro-tumorigenic systemic environment is established.     

Reprogramming of human cancer cells to pluripotency for models of cancer progression

The ability to study live cells as they progress through the stages of cancer provides the opportunity to discover dynamic networks underlying pathology, markers of early stages, and ways to assess therapeutics. Genetically engineered animal models of cancer, where it is possible to study the consequences of temporal‐specific induction of oncogenes or deletion of tumor suppressors, have yielded major insights into cancer progression. Yet differences exist between animal and human cancers, such as in markers of progression and response to therapeutics. Thus, there is a need for human cell models of cancer progression. Most human cell models of cancer are based on tumor cell lines and xenografts of primary tumor cells that resemble the advanced tumor state, from which the cells were derived, and thus do not recapitulate disease progression. Yet a subset of cancer types have been reprogrammed to pluripotency or near‐pluripotency by blastocyst injection, by somatic cell nuclear transfer and by induced pluripotent stem cell (iPS) technology. The reprogrammed cancer cells show that pluripotency can transiently dominate over the cancer phenotype. Diverse studies show that reprogrammed cancer cells can, in some cases, exhibit early‐stage phenotypes reflective of only partial expression of the cancer genome. In one case, reprogrammed human pancreatic cancer cells have been shown to recapitulate stages of cancer progression, from early to late stages, thus providing a model for studying pancreatic cancer development in human cells where previously such could only be discerned from mouse models. We discuss these findings, the challenges in developing such models and their current limitations, and ways that iPS reprogramming may be enhanced to develop human cell models of cancer progression.     

The role of focal adhesion kinase in tumor initiation and progression

Focal adhesion kinase (FAK) is a nonreceptor tyrosine kinase that acts as a primary regulator of focal adhesion signaling to regulate cell proliferation, survival, and migration. While FAK is known to directly influence many fundamental adhesion and growth factor signaling pathways important in cancer, and FAK is overexpressed in multiple human cancers, studies addressing a causal role for FAK in tumor initiation and progression using transgenic models of human cancer had not been performed. Recently, using tissue-specific FAK-knockout in mouse models of human cancer, the consequences of FAK ablation in carcinoma were demonstrated by multiple independent research groups. Strong consensus evidence indicates that epithelial cells are able to transform in the absence of FAK, but do not undergo a malignant conversion to invasive carcinoma, and as such, metastasis is impaired. This is likely the consequence of decreased Src and p130Cas activation in concert with misregulated actin cytoskeleton dynamics and Rho GTPase signaling. Hence, FAK, as well as the FAK-regulating/regulated signaling network, are viable candidates for cancer metastasis therapies.     

Rtreemix: an R package for estimating evolutionary pathways and genetic progression scores

In genetics, many evolutionary pathways can be modeled by the ordered accumulation of permanent changes. Mixture models of mutagenetic trees have been used to describe disease progression in cancer and in HIV. In cancer, progression is modeled by the accumulation of chromosomal gains and losses in tumor cells; in HIV, the accumulation of drug resistance-associated mutations in the viral genome is known to be associated with disease progression. From such evolutionary models, genetic progression scores can be derived that assign measures for the disease state to single patients. Rtreemix is an R package for estimating mixture models of evolutionary pathways from observed cross-sectional data and for estimating associated genetic progression scores. The package also provides extended functionality for estimating confidence intervals for estimated model parameters and for evaluating the stability of the estimated evolutionary mixture models.