Statistical behavior of complex cancer karyotypes

Overexpression of HES6 has prognostic value and promotes metastasis via the Wnt/β-catenin signaling pathway in colorectal cancer

HES6 is a member of the hairy-enhancer of the split homolog family, which has been implicated in oncogenesis and cancer progression in a variety of human cancers, including prostate and breast cancer. However, its clinical significance and biological role in colorectal cancer (CRC) remain unclear. In the present study, the expression of HES6 was significantly upregulated in CRC cell lines and CRC tissues at both the mRNA and protein levels. The present study also reported high expression of HES6 in 138/213 (64.8%) paraffin-embedded archived CRC specimens. HES6 expression was significantly correlated with T classification (P<0.001), N classification (P=0.020), and distant metastasis (P<0.001). Patients with higher HES6 expression levels exhibited a reduced overall survival (P<0.001). In addition, a multivariate analysis revealed that the expression of HES6 may be a novel prognostic marker for the survival of patients with CRC. Furthermore, the present study demonstrated that ectopic expression of HES6 enhanced the migration and invasive abilities of CRC cells. These abilities were significantly inhibited upon knockdown of endogenous HES6 expression by specific short hairpin RNAs. Additionally, the present study reported that the effects of HES6 on metastasis may be associated with the activation of the Wnt/β-catenin signaling pathway. Collectively, the findings of the present study revealed that overexpression of HES6 played a key role in the progression of CRC, leading to a poor prognosis and clinical outcome.

Characterization of human gastric adenocarcinoma cell lines established from peritoneal ascites

The three cell lines, designated as gastric cancer (GC)1401, GC1415 and GC1436 were derived from peritoneal effusions from patients with gastric adenocarcinoma. Cell lines were established in tissue culture and in immunodeficient, non-obese diabetic/severe combined immunodeficiency (NOD/SCID) mice. All cell lines were cultured in Dulbecco's modified Eagle's medium supplemented with 5% fetal bovine serum. These cell lines were grown as an adherent monolayer with doubling time ranging between 25 h (GC1436 cell line) and 30–34 h (GC1401 and GC1415, respectively). All cells showed morphological features of epithelial-like cells, forming sheets of polygonal cells. Chromosomal analysis showed that the modal numbers ranged from 52 (GC1401), 51–56 (GC1415) and 106 (GC1436). High heterogeneity, resulting from several structural and numerical chromosomal abnormalities were evident in all cell lines. The surface marker expression suggested a tumor origin of the cells, and indicated the intestinal phenotype of a GC (CD10⁺, MUC1). All three cell lines were tumorigenic but not metastatic, in vivo, in NOD/SCID mice. The lack of metastatic potential was suggested by the lack of aldehyde dehydrogenase 1A1 activity. In conclusion, these newly established GC cell lines widen the feasibility of the functional studies on biology of GC as well as drug testing for potential therapeutic purposes.

Different Expression and Localization of Phosphoinositide Specific Phospholipases C in Human Osteoblasts, Osteosarcoma Cell Lines, Ewing Sarcoma and Synovial Sarcoma

Background: Bone hardness and strength depends on mineralization, which involves a complex process in which calcium phosphate, produced by bone-forming cells, was shed around the fibrous matrix. This process is strictly regulated, and a number of signal transduction systems were interested in calcium metabolism, such as the phosphoinositide (PI) pathway and related phospholipase C (PLC) enzymes. Objectives: Our aim was to search for common patterns of expression in osteoblasts, as well as in ES and SS. Methods: We analysed the PLC enzymes in human osteoblasts and osteosarcoma cell lines MG-63 and SaOS-2. We compared the obtained results to the expression of PLCs in samples of patients affected with Ewing sarcoma (ES) and synovial sarcoma (SS). Results: In osteoblasts, MG-63 cells and SaOS-2 significant differences were identified in the expression of PLC δ4 and PLC η subfamily isoforms. Differences were also identified regarding the expression of PLCs in ES and SS. Most ES and SS did not express PLCB1, which was expressed in most osteoblasts, MG-63 and SaOS-2 cells. Conversely, PLCB2, unexpressed in the cell lines, was expressed in some ES and SS. However, PLCH1 was expressed in SaOS-2 and inconstantly expressed in osteoblasts, while it was expressed in ES and unexpressed in SS. The most relevant difference observed in ES compared to SS regarded PLC ε and PLC η isoforms. Conclusion: MG-63 and SaOS-2 osteosarcoma cell lines might represent an inappropriate experimental model for studies about the analysis of signal transduction in osteoblasts

The mathematics of cancer: integrating quantitative models

Mathematical modelling approaches have become increasingly abundant in cancer research. The complexity of cancer is well suited to quantitative approaches as it provides challenges and opportunities for new developments. In turn, mathematical modelling contributes to cancer research by helping to elucidate mechanisms and by providing quantitative predictions that can be validated. The recent expansion of quantitative models addresses many questions regarding tumour initiation, progression and metastases as well as intra-tumour heterogeneity, treatment responses and resistance. Mathematical models can complement experimental and clinical studies, but also challenge current paradigms, redefine our understanding of mechanisms driving tumorigenesis and shape future research in cancer biology.

Can a minimal replicating construct be identified as the embodiment of cancer?

Genomic instability is a hallmark of cancer. Cancer cells that exhibit abnormal chromosomes are characteristic of most advanced tumours, despite the potential threat represented by accumulated genetic damage. Carcinogenesis involves a loss of key components of the genetic and signalling molecular networks; hence some authors have suggested that this is part of a trend of cancer cells to behave as simple, minimal replicators. In this study, we explore this conjecture and suggest that, in the case of cancer, genomic instability has an upper limit that is associated with a minimal cancer cell network. Such a network would include (for a given microenvironment) the basic molecular components that allow cells to replicate and respond to selective pressures. However, it would also exhibit internal fragilities that could be exploited by appropriate therapies targeting the DNA repair machinery. The implications of this hypothesis are discussed.

Inferring mutational timing and reconstructing tumour evolutionary histories

Cancer evolution can be considered within a Darwinian framework. Both micro and macro-evolutionary theories can be applied to understand tumour progression and treatment failure. Owing to cancers' complexity and heterogeneity the rules of tumour evolution, such as the role of selection, remain incompletely understood. The timing of mutational events during tumour evolution presents diagnostic, prognostic and therapeutic opportunities. Here we review the current sampling and computational approaches for inferring mutational timing and the evidence from next generation sequencing-informed data on mutational timing across all tumour types. We discuss how this knowledge can be used to illuminate the genes and pathways that drive cancer initiation and relapse; and to support drug development and clinical trial design.

Sequential and hierarchical chromosomal changes and chromosome instability are distinct features of high hyperdiploid pediatric acute lymphoblastic leukemia

BACKGROUND

Pathogenesis of the non-random accumulation of extra chromosomes in the low and high hyperdiploid (HeL, HeH) pre-B pediatric acute lymphoblastic leukemia (B-pALL) is largely unknown, and has been clarified with respect only to tetrasomic chromosomes. We analyzed the hierarchy of changes in chromosome number and chromosomal instability, as well as clonal heterogeneity and evolution, in the untreated bone marrow cell samples from 214 B-pALL patients.

PROCEDURE

Applying relocation, 2 × 4 color interphase fluorescence in situ hybridization was used to detect copy number alterations (CNAs) of the most commonly involved chromosomes, 4, 6, 10, 14, 17, 18, 21, and X. This approach allowed us to acquire a dataset correlated for all eight parameters.

RESULTS

Based on chromosome number, an average of 6.9 and 10.2, whereas according to unique constellation 15.3 and 26.7 subclones could be identified in the HeL and HeH subgroups, respectively. Cluster analysis revealed the order of CNAs to chromosomes was highly conserved, and network analysis indicated changes in chromosome number were sequential for 80-90% of all numerical aberrations. Significant chromosome instability was revealed in both subgroups of leukemia.

CONCLUSIONS

Data generated using this new approach indicate that chromosomal instability, which causes heterogeneity in the subclonal landscape, and the sequential changes to chromosome numbers, are both determining factors in the pathomechanism of the hyperdiploid B-pALL. These new observations could prompt research into the mitotic machinery of leukemic cells to identify new therapeutic targets for treating this disease.

Ovarian cancer evolution through stochastic genome alterations: defining the genomic role in ovarian cancer

Ovarian cancer is the fifth leading cause of death among women worldwide. Characterized by complex etiology and multi-level heterogeneity, its origins are not well understood. Intense research efforts over the last decade have furthered our knowledge by identifying multiple risk factors that are associated with the disease. However, it is still unclear how genetic heterogeneity contributes to tumor formation, and more specifically, how genome-level heterogeneity acts as the key driving force of cancer evolution. Most current genomic approaches are based on 'average molecular profiling.' While effective for data generation, they often fail to effectively address the issue of high level heterogeneity because they mask variation that exists in a cell population. In this synthesis, we hypothesize that genome-mediated cancer evolution can effectively explain diverse factors that contribute to ovarian cancer. In particular, the key contribution of genome replacement can be observed during major transitions of ovarian cancer evolution including cellular immortalization, transformation, and malignancy. First, we briefly review major updates in the literature, and illustrate how current gene-mediated research will offer limited insight into cellular heterogeneity and ovarian cancer evolution. We next explain a holistic framework for genome-based ovarian cancer evolution and apply it to understand the genomic dynamics of a syngeneic ovarian cancer mouse model. Finally, we employ single cell assays to further test our hypothesis, discuss some predictions, and report some recent findings.

Elevated tolerance to aneuploidy in cancer cells: estimating the fitness effects of chromosome number alterations by in silico modelling of somatic genome evolution

An unbalanced chromosome number (aneuploidy) is present in most malignant tumours and has been attributed to mitotic mis-segregation of chromosomes. However, recent studies have shown a relatively high rate of chromosomal mis-segregation also in non-neoplastic human cells, while the frequency of aneuploid cells remains low throughout life in most normal tissues. This implies that newly formed aneuploid cells are subject to negative selection in healthy tissues and that attenuation of this selection could contribute to aneuploidy in cancer. To test this, we modelled cellular growth as discrete time branching processes, during which chromosome gains and losses were generated and their host cells subjected to selection pressures of various magnitudes. We then assessed experimentally the frequency of chromosomal mis-segregation as well as the prevalence of aneuploid cells in human non-neoplastic cells and in cancer cells. Integrating these data into our models allowed estimation of the fitness reduction resulting from a single chromosome copy number change to an average of ≈30% in normal cells. In comparison, cancer cells showed an average fitness reduction of only 6% (p = 0.0008), indicative of aneuploidy tolerance. Simulations based on the combined presence of chromosomal mis-segregation and aneuploidy tolerance reproduced distributions of chromosome aberrations in >400 cancer cases with higher fidelity than models based on chromosomal mis-segregation alone. Reverse engineering of aneuploid cancer cell development in silico predicted that aneuploidy intolerance is a stronger limiting factor for clonal expansion of aneuploid cells than chromosomal mis-segregation rate. In conclusion, our findings indicate that not only an elevated chromosomal mis-segregation rate, but also a generalised tolerance to novel chromosomal imbalances contribute to the genomic landscape of human tumours.

Dissecting karyotypic patterns in non-hyperdiploid multiple myeloma: an overview on the karyotypic evolution

BACKGROUND

Multiple myeloma (MM) is a plasma cell disorder characterized by the presence of specific genetic and cytogenetic aberrations that define unique subgroups with different outcomes. On the basis of the ploidy status, MM can be subdivided into hyperdiploid MM (H-MM) and non-hyperdiploid MM (NH-MM). NH-MM is an entity that encompasses hypodiploid, pseudodiploid, and near tetraploid MM and is associated with a higher number of immunoglobulin heavy-chain (IgH) translocations.

MATERIALS AND METHODS

We have systematically analyzed the structure of the karyotypic evolution in NH-MM and identified several genetic features of their complex karyotypic patterns.

RESULTS

On the basis of statistical models used in complex karyotypes, we were able to identify the temporal order in which the genetic aberrations occur in NH-MM. In this analysis, whole chromosome losses and IgH translocations were commonly seen, and -13/13q- and t14q32 were defined as early genetic events in the karyotypic evolution of NH-MM. Furthermore, chromosome 1 and 17 abnormalities were associated with a late karyotypic phase of evolution consistent with the recognized pattern of acquired events deemed to be associated with these type of genetic aberrations.

CONCLUSION

Accumulation of genetic aberrations in NH-MM above a threshold results in malignant transformation.

Genotyping techniques to address diversity in tumors

Array-based genotyping platforms have during recent years been established as a valuable tool for the characterization of genomic alterations in cancer. The analysis of tumor samples, however, presents challenges for data analysis and interpretation. For example, tumor samples are often admixed with nonaberrant cells that define the tumor microenvironment, such as infiltrating lymphocytes and fibroblasts, or vasculature. Furthermore, tumors often comprise subclones harboring divergent aberrations that are acquired subsequent to the tumor-initiating event. The combined analysis of both genotype and copy number status obtained by array-based genotyping platforms provide opportunities to address these challenges. In this chapter, we present the basic principles for current array-based genotyping platforms and how they can be used to infer genotype and copy number for acquired genomic alterations. We describe how these techniques can be used to resolve tumor ploidy, normal cell admixture, and subclonality. We also exemplify how genotyping techniques can be applied in tumor studies to elucidate the hierarchy among tumor clones, and thus, provide means to study clonal expansion and tumor evolution.

A mathematical methodology for determining the temporal order of pathway alterations arising during gliomagenesis

Human cancer is caused by the accumulation of genetic alterations in cells. Of special importance are changes that occur early during malignant transformation because they may result in oncogene addiction and thus represent promising targets for therapeutic intervention. We have previously described a computational approach, called Retracing the Evolutionary Steps in Cancer (RESIC), to determine the temporal sequence of genetic alterations during tumorigenesis from cross-sectional genomic data of tumors at their fully transformed stage. Since alterations within a set of genes belonging to a particular signaling pathway may have similar or equivalent effects, we applied a pathway-based systems biology approach to the RESIC methodology. This method was used to determine whether alterations of specific pathways develop early or late during malignant transformation. When applied to primary glioblastoma (GBM) copy number data from The Cancer Genome Atlas (TCGA) project, RESIC identified a temporal order of pathway alterations consistent with the order of events in secondary GBMs. We then further subdivided the samples into the four main GBM subtypes and determined the relative contributions of each subtype to the overall results: we found that the overall ordering applied for the proneural subtype but differed for mesenchymal samples. The temporal sequence of events could not be identified for neural and classical subtypes, possibly due to a limited number of samples. Moreover, for samples of the proneural subtype, we detected two distinct temporal sequences of events: (i) RAS pathway activation was followed by TP53 inactivation and finally PI3K2 activation, and (ii) RAS activation preceded only AKT activation. This extension of the RESIC methodology provides an evolutionary mathematical approach to identify the temporal sequence of pathway changes driving tumorigenesis and may be useful in guiding the understanding of signaling rearrangements in cancer development.

Cancer is a deadly disease that develops through the accumulation of genetic changes over time. Many biological models do not incorporate this temporal aspect of tumor formation and progression, in part due to the difficulty of determining the sequence of events through biological experimentation for most cancer types. We previously developed a computational algorithm with which we can quickly and cost-effectively determine the order in which mutations arise in the tumor even when large numbers of mutations are considered. In this paper, we extended our method to incorporate biological knowledge of the common pathways by which cancer progresses. We applied these techniques to primary glioblastoma, the most common form of brain cancer. We found that when all samples are taken into account, a temporal sequence of pathway events emerges; however, different subtypes of glioblastoma vary in their temporal sequence of events. This algorithm can also be easily applied to other cancer types as clinical data becomes available, showing the benefit of computational and mathematical tools in cancer research. Using temporal information, cancer biologists will be able to develop more accurate animal models of tumor formation and learn more about how mutations interact in time, thus leading to better treatments for cancer.

Evolutionary mechanisms and diversity in cancer

The recently introduced genome theory of cancer evolution provides a new framework for evolutionary studies on cancer. In particular, the established relationship between the large number of individual molecular mechanisms and the general evolutionary mechanism of cancer calls upon a change in our strategies that have been based on the characterization of common cancer gene mutations and their defined pathways. To further explain the significance of the genome theory of cancer evolution, a brief review will be presented describing the various attempts to illustrate the evolutionary mechanism of cancer, followed by further analysis of some key components of somatic cell evolution, including the diversity of biological systems, the multiple levels of information systems and control systems, the two phases (the punctuated or discontinuous phase and gradual Darwinian stepwise phase) and dynamic patterns of somatic cell evolution where genome replacement is the driving force. By linking various individual molecular mechanisms to the level of genome population diversity and tumorigenicity, the general mechanism of cancer has been identified as the evolutionary mechanism of cancer, which can be summarized by the following three steps including stress-induced genome instability, population diversity or heterogeneity, and genome-mediated macroevolution. Interestingly, the evolutionary mechanism is equal to the collective aggregate of all individual molecular mechanisms. This relationship explains why most of the known molecular mechanisms can contribute to cancer yet there is no single dominant mechanism for the majority of clinical cases. Despite the fact that each molecular mechanism can serve as a system stress and initiate the evolutionary process, to achieve cancer, multiple cycles of genome-mediated macroevolution are required and are a stochastically determined event. Finally, the potential clinical implications of the evolutionary mechanism of cancer are briefly reviewed.

Large-scale analysis of chromosomal aberrations in cancer karyotypes reveals two distinct paths to aneuploidy

Background

Chromosomal aneuploidy, that is to say the gain or loss of chromosomes, is the most common abnormality in cancer. While certain aberrations, most commonly translocations, are known to be strongly associated with specific cancers and contribute to their formation, most aberrations appear to be non-specific and arbitrary, and do not have a clear effect. The understanding of chromosomal aneuploidy and its role in tumorigenesis is a fundamental open problem in cancer biology.

Results

We report on a systematic study of the characteristics of chromosomal aberrations in cancers, using over 15,000 karyotypes and 62 cancer classes in the Mitelman Database. Remarkably, we discovered a very high co-occurrence rate of chromosome gains with other chromosome gains, and of losses with losses. Gains and losses rarely show significant co-occurrence. This finding was consistent across cancer classes and was confirmed on an independent comparative genomic hybridization dataset of cancer samples. The results of our analysis are available for further investigation via an accompanying website.

Conclusions

The broad generality and the intricate characteristics of the dichotomy of aneuploidy, ranging across numerous tumor classes, are revealed here rigorously for the first time using statistical analyses of large-scale datasets. Our finding suggests that aneuploid cancer cells may use extra chromosome gain or loss events to restore a balance in their altered protein ratios, needed for maintaining their cellular fitness.

CDCOCA: a statistical method to define complexity dependence of co-occuring chromosomal aberrations

Background

Copy number alterations (CNA) play a key role in cancer development and progression. Since more than one CNA can be detected in most tumors, frequently co-occurring genetic CNA may point to cooperating cancer related genes. Existing methods for co-occurrence evaluation so far have not considered the overall heterogeneity of CNA per tumor, resulting in a preferential detection of frequent changes with limited specificity for each association due to the high genetic instability of many samples.

Method

We hypothesize that in cancer some linkage-independent CNA may display a non-random co-occurrence, and that these CNA could be of pathogenetic relevance for the respective cancer. We also hypothesize that the statistical relevance of co-occurring CNA may depend on the sample specific CNA complexity. We verify our hypotheses with a simulation based algorithm CDCOCA (complexity dependence of co-occurring chromosomal aberrations).

Results

Application of CDCOCA to example data sets identified co-occurring CNA from low complex background which otherwise went unnoticed. Identification of cancer associated genes in these co-occurring changes can provide insights of cooperative genes involved in oncogenesis.

Conclusions

We have developed a method to detect associations of regional copy number abnormalities in cancer data. Along with finding statistically relevant CNA co-occurrences, our algorithm points towards a generally low specificity for co-occurrence of regional imbalances in CNA rich samples, which may have negative impact on pathway modeling approaches relying on frequent CNA events.

Statistical analysis of the cancer cell's molecular entropy using high-throughput data

MOTIVATION

As cancer progresses, DNA copy number aberrations accumulate and the genomic entropy (chromosomal disorganization) increases. For this surge to have any oncogenetic effect, it should (to some extent) be reflected at other molecular levels of the cancer cell, in particular that of the transcriptome. Such a coincidence of cancer progression and the propagation of an entropy increase through the molecular levels of the cancer cell would enhance the understanding of cancer evolution.

RESULTS

A statistical argument reveals that (under some assumptions) an entropy increase in one random variable (DNA copy number) leads to an entropy increase in another (gene expression). Statistical methodology is provided to investigate the relation between the genomic and transcriptomic entropy using high-throughput data. Analyses of multiple high-throughput datasets using this methodology show a close, concordant relation among the genomic and transcriptomic entropy. Hence, as cancer evolves, and the genomic entropy increases, the transcriptomic entropy is also expected to surge.

Frequency of double minute chromosomes and combined cytogenetic abnormalities and their characteristics

Double minute chromosomes (DMs) are the cytogenetic hallmark of extra-chromosomal genomic amplification. The frequency of DMs in primary cancer and the cytogenetic features of DMs-positive primary cancer cases are largely unknown. To unravel these issues, we retrieved the Mitelman database and analyzed all DMs-positive primary cancerous karyotypes (787 karyotypes). The overall frequency of DMs is 1.4% (787 DMs-positive cases; total 54,398 cases). We found that DMs have the highest frequency in adrenal carcinoma (28.6%, topography) and neuroblastoma (31.7%, morphology). The frequencies of DMs in each tumor were much lower than in previous reports. The frequency of DMs in malignant cancers is significantly higher than in benign cancers, which confirms that DMs are malignant cytogenetic markers. DMs combined cytogenetic abnormalities are identified and sorted into two groups by principal component analysis (PCA), with one group containing -4, -5, -8, -9, -10, -13, -14, -15, -16, -17, -18, -20, -21, and -22, and the other containing -1p, -5q, +7, and +20. The prominent imbalance in DMs-positive cancer cases is chromosome loss. However, DMs-positive cancer cases, deriving from different morphologic cancers, cannot be clearly divided into subgroups. Our large database analysis provides novel knowledge of DMs and their combined cytogenetic abnormalities in primary cancer.

Bladder cancer or bladder cancers? Genetically distinct malignant conditions of the urothelium

Despite the fact that the current histopathologic classification for bladder cancer has led to improved concepts for the clinical management of the disease, key questions with regard to assessment of risk for recurrence and/or progression to invasive disease remain. In addition, response to specific therapies cannot be predicted accurately. Bladder tumors comprise a heterogeneous group with respect to both histopathology and clinical behavior. Thus, it is anticipated that a thorough knowledge and interpretation of the molecular alterations involved in tumor development and progression will lead to greater prognostic and predictive power. This may not only lead to better comprehension of the biology of the disease, but may also lead to the development of novel individualized therapies. Novel means of stratification are urgently needed to provide a new subclassification of urothelial lesions. This review discusses and summarizes the genetic alterations that have been reported in bladder cancer and relates these to the current 2-pathway model for tumor development. The molecular pathogenesis of high-grade noninvasive papillary tumors and of T1 tumors is not yet clear, and possibilities are discussed.

Tracing the tumor lineage

Defining the pathways through which tumors progress is critical to our understanding and treatment of cancer. We do not routinely sample patients at multiple time points during the progression of their disease, and thus our research is limited to inferring progression a posteriori from the examination of a single tumor sample. Despite this limitation, inferring progression is possible because the tumor genome contains a natural history of the mutations that occur during the formation of the tumor mass. There are two approaches to reconstructing a lineage of progression: (1) inter-tumor comparisons, and (2) intra-tumor comparisons. The inter-tumor approach consists of taking single samples from large collections of tumors and comparing the complexity of the genomes to identify early and late mutations. The intra-tumor approach involves taking multiple samples from individual heterogeneous tumors to compare divergent clones and reconstruct a phylogenetic lineage. Here we discuss how these approaches can be used to interpret the current models for tumor progression. We also compare data from primary and metastatic copy number profiles to shed light on the final steps of breast cancer progression. Finally, we discuss how recent technical advances in single cell genomics will herald a new era in understanding the fundamental basis of tumor heterogeneity and progression.

Sorting cancer karyotypes by elementary operations

Since the discovery of the "Philadelphia chromosome" in chronic myelogenous leukemia in 1960, there has been ongoing intensive research of chromosomal aberrations in cancer. These aberrations, which result in abnormally structured genomes, became a hallmark of cancer. Many studies provide evidence for the connection between chromosomal alterations and aberrant genes involved in the carcinogenesis process. An important problem in the analysis of cancer genomes is inferring the history of events leading to the observed aberrations. Cancer genomes are usually described in the form of karyotypes, which present the global changes in the genomes' structure. In this study, we propose a mathematical framework for analyzing chromosomal aberrations in cancer karyotypes. We introduce the problem of sorting karyotypes by elementary operations, which seeks a shortest sequence of elementary chromosomal events transforming a normal karyotype into a given (abnormal) cancerous karyotype. Under certain assumptions, we prove a lower bound for the elementary distance, and present a polynomial-time 3-approximation algorithm for the problem. We applied our algorithm to karyotypes from the Mitelman database, which records cancer karyotypes reported in the scientific literature. Approximately 94% of the karyotypes in the database, totaling 58,464 karyotypes, supported our assumptions, and each of them was subjected to our algorithm. Remarkably, even though the algorithm is only guaranteed to generate a 3-approximation, it produced a sequence whose length matched the lower bound (and hence optimal) in 99.9% of the tested karyotypes.

Quantifying cancer progression with conjunctive Bayesian networks

MOTIVATION

Cancer is an evolutionary process characterized by accumulating mutations. However, the precise timing and the order of genetic alterations that drive tumor progression remain enigmatic.

RESULTS

We present a specific probabilistic graphical model for the accumulation of mutations and their interdependencies. The Bayesian network models cancer progression by an explicit unobservable accumulation process in time that is separated from the observable but error-prone detection of mutations. Model parameters are estimated by an Expectation-Maximization algorithm and the underlying interaction graph is obtained by a simulated annealing procedure. Applying this method to cytogenetic data for different cancer types, we find multiple complex oncogenetic pathways deviating substantially from simplified models, such as linear pathways or trees. We further demonstrate how the inferred progression dynamics can be used to improve genetics-based survival predictions which could support diagnostics and prognosis.

AVAILABILITY

The software package ct-cbn is available under a GPL license on the web site cbg.ethz.ch/software/ct-cbn

CONTACT

moritz.gerstung@bsse.ethz.ch.

Inferring progression models for CGH data

MOTIVATION

One of the mutational processes that has been monitored genome-wide is the occurrence of regional DNA copy number alterations (CNAs), which may lead to deletion or over-expression of tumor suppressors or oncogenes, respectively. Understanding the relationship between CNAs and different cancer types is a fundamental problem in cancer studies.

RESULTS

This article develops an efficient method that can accurately model the progression of the cancer markers and reconstruct evolutionary relationship between multiple types of cancers using comparative genomic hybridization (CGH) data. Such modeling can lead to better understanding of the commonalities and differences between multiple cancer types and potential therapies. We have developed an automatic method to infer a graph model for the markers of multiple cancers from a large population of CGH data. Our method identifies highly related markers across different cancer types. It then builds a directed acyclic graph that shows the evolutionary history of these markers based on how common each marker is in different cancer types. We demonstrated the use of this model in determining the importance of markers in cancer evolution. We have also developed a new method to measure the evolutionary distance between different cancers based on their markers. This method employs the graph model we developed for the individual markers to measure the distance between pairs of cancers. We used this measure to create an evolutionary tree for multiple cancers. Our experiments on Progenetix database show that our markers are largely consistent to the reported hot-spot imbalances and most frequent imbalances. The results show that our distance measure can accurately reconstruct the evolutionary relationship between multiple cancer types.

Unusually stable abnormal karyotype in a highly aggressive melanoma negative for telomerase activity

Malignant melanomas are characterized by increased karyotypic complexity, extended aneuploidy and heteroploidy. We report a melanoma metastasis to the peritoneal cavity with an exceptionally stable, abnormal pseudodiploid karyotype as verified by G-Banding, subtelomeric, centromeric and quantitative Fluorescence in Situ Hybridization (FISH). Interestingly this tumor had no detectable telomerase activity as indicated by the Telomere Repeat Amplification Protocol. Telomeric Flow-FISH and quantitative telomeric FISH on mitotic preparations showed that malignant cells had relatively short telomeres. Microsatellite instability was ruled out by the allelic pattern of two major mononucleotide repeats. Our data suggest that a combination of melanoma specific genomic imbalances were sufficient and enough for this fatal tumor progression, that was not accompanied by genomic instability, telomerase activity, or the engagement of the alternative recombinatorial telomere lengthening pathway.

Molecular pathogenesis of bladder cancer

Bladder tumors show widely differing histopathology and clinical behavior. This is reflected in the molecular genetic alterations they contain. Much information has accumulated on somatic genomic alterations in bladder tumors of all grades and stages and when this information is related to the common histopathological appearances, a model for the pathogenesis of two major groups of bladder tumors has emerged. This review summarizes the genetic alterations that have been reported in bladder cancer and relates these to the current two-pathway model for tumor development. The molecular pathogenesis of high-grade noninvasive papillary tumors and of T1 tumors is not yet clear and possibilities are discussed.

Classification and feature selection algorithms for multi-class CGH data

Recurrent chromosomal alterations provide cytological and molecular positions for the diagnosis and prognosis of cancer. Comparative genomic hybridization (CGH) has been useful in understanding these alterations in cancerous cells. CGH datasets consist of samples that are represented by large dimensional arrays of intervals. Each sample consists of long runs of intervals with losses and gains. In this article, we develop novel SVM-based methods for classification and feature selection of CGH data. For classification, we developed a novel similarity kernel that is shown to be more effective than the standard linear kernel used in SVM. For feature selection, we propose a novel method based on the new kernel that iteratively selects features that provides the maximum benefit for classification. We compared our methods against the best wrapper-based and filter-based approaches that have been used for feature selection of large dimensional biological data. Our results on datasets generated from the Progenetix database, suggests that our methods are considerably superior to existing methods.

AVAILABILITY

All software developed in this article can be downloaded from http://plaza.ufl.edu/junliu/feature.tar.gz.

Establishment and molecular cytogenetic characterization of non-small cell lung cancer cell line KU-T1 by multicolor fluorescence in situ hybridization, comparative genomic hybridization, and chromosome microdissection

A human lung adenocarcinoma cell line, designated KU-T1, was established from a Japanese man in Kochi Medical School. Conventional banding and multicolor fluorescence in situ hybridization (M-FISH) analyses of KU-T1 cells revealed a hyperdiploid chromosomal constitution and complex karyotypes. Comparative genomic hybridization showed several chromosomal copy number changes, and five regions that were highly amplified. Two of the five highly amplified regions, 1q and 3q, were identified from distributions of DNA sequences on a metaphase cell by FISH using chromosome microdissection-generated probes hybridized to 1q32 approximately q34 and 3q26 approximately q28, respectively. The 3q probe depicted a homogeneously staining region (hsr) in a derivative chromosome 3 of KU-T1. An hsr probe was regenerated by chromosome microdissection and was hybridized back to KU-T1 and normal metaphases. This hybridization experiment confirmed the probe derived from an hsr and indicated original locations of DNA sequences of hsr on normal chromosome 3. Intense hybridized signals shown at three loci (3p12, 3q26.3, and 3q28) suggests that oncogenes may be involved in the hsr formation. The present study provides a comprehensive analysis of the chromosomal abnormalities, including hsr formation and related oncogenes, in the KU-T1 cell line.

The coincidence of chromosome 15 aberrations and beta2-microglobulin gene mutations is causative for the total loss of human leukocyte antigen class I expression in melanoma

PURPOSE

Total loss of surface presentation of human leukocyte antigen (HLA) class I molecules, protecting tumor cells from the recognition by cytotoxic host CD8+ T cells, is known to be caused by mutations in the beta2-microglobulin (beta2m) gene. We asked whether abnormalities of chromosome 15, harboring the beta2m gene on 15q21, in addition to beta2m gene mutations, are causative for the HLA class I-negative phenotype of melanoma cells.

EXPERIMENTAL DESIGN

To answer this, we established primary cell lines from the beta2m-negative metastatic melanoma tissues of four different patients and analyzed them for beta2m gene mutations and chromosome 15 aberrations, the latter by loss of heterozygosity analysis, fluorescence in situ hybridization (FISH), and multicolor FISH.

RESULTS

Mutations at the beta2m gene level were detected in all cell lines. The loss of heterozygosity analysis of microsatellite markers located on chromosome 15 in three of the four cell lines pointed to an extensive loss of chromosome 15 material. Subsequent molecular cytogenetic analysis revealed the coexistence of apparently normal and rearranged versions of chromosome 15 in three cell lines whereas the fourth cell line solely showed rearranged versions. Two of the four cell lines exhibited a special type of intrachromosomal rearrangement characterized by FISH signals specific for the subtelomeric region of 15q at both ends of the chromosome and one centromeric signal in between.

CONCLUSIONS

Our data indicate that the complete loss of HLA class I expression in melanoma cells is due to the coincidence of the following mutational events: (a) chromosome 15 instability associated with an extensive loss of genetic material and (b) beta2m gene mutations.

Chromosomal abnormalities of adenocarcinoma of the pancreas: identifying early and late changes

The high level of karyotypic complexity found in epithelial neoplasms hinders the characterization of their cytogenetic evolution. Derivation of such pathways in adenocarcinoma of the pancreas has been particularly limited, because only a few pancreatic carcinomas are resected at an early stage of disease and so the number of primary carcinomas for which analysis of abnormal karyotypes has been reported is small. Here we report the clonal karyotypic abnormalities identified by G-banding analysis of 36 primary pancreatic carcinomas obtained from patients undergoing a Whipple resection with curative intent. The majority of the 36 carcinomas were diploid or triploid (33 of 36; 91%). Numerical alterations were found in all carcinomas for which a complete karyotype was determined. All the chromosomes were involved in gain, loss, or both gain and loss of the entire chromosome, in at least 8 and up to 28 of the carcinomas. Most commonly lost were chromosomes 18 (in 78% of the 36 carcinomas), 17 (56%), 6 (44%), 21 (42%), 22 (42%), Y (36%), and 4 (33%). Gain of chromosome 20 was observed in 10 of the 36 carcinomas. Structural abnormalities were common, resulting in partial chromosomal gains and losses, with a median number of 7 partial imbalances per carcinoma (range, 1-15). Sixteen carcinomas contained double-minute chromosomes, homogeneously staining regions, or both, indicating gene amplification. Pooling data for these 36 carcinomas with the primary carcinomas with karyotypes published in the Mitelman database (http://cgap.nci.nih.gov/Chromosomes/Mitelman), we defined pathways of karyotypic evolution. The most frequent chromosomal imbalances were -18 (67.6%), -10 (34.3%), -4 (31.4%), +20 (31.4%), -15p (23.8%), -14p (22.9%), +2 (21.9%), -5 (21.9%), -13p (20%), +16 (20%), -21p (19%), -17p (19%), +1q (19.0%). Recurrent imbalances identified as occurring early were -1p, -15p, -18, -7q, -8p, -17p, and -5; late recurrent imbalances were +11q, +7q, +6p, -19p, and +2. In contrast to reports from similar analyses in other epithelial carcinomas, we did not find evidence for multiple karyotypic evolutionary pathways.

Distinct evolutionary mechanisms for genomic imbalances in high-risk and low-risk neuroblastomas

BACKGROUND

Neuroblastoma (NB) is the most common extracranial solid tumour of childhood. Several genomic imbalances correlate to prognosis in NB, with structural rearrangements, including gene amplification, in a near-diploid setting typically signifying high-risk tumours and numerical changes in a near-triploid setting signifying low-risk tumours. Little is known about the temporal sequence in which these imbalances occur during the carcinogenic process.

METHODS

We have reconstructed the appearance of cytogenetic imbalances in 270 NBs by first grouping tumours and imbalances through principal component analysis and then using the number of imbalances in each tumour as an indicator of evolutionary progression.

RESULTS

Tumours clustered in four sub-groups, dominated respectively by (1) gene amplification in double minute chromosomes and few other aberrations, (2) gene amplification and loss of 1p sequences, (3) loss of 1p and other structural aberrations including gain of 17q, and (4) whole-chromosome gains and losses. Temporal analysis showed that the structural changes in groups 1-3 were acquired in a step-wise fashion, with loss of 1p sequences and the emergence of double minute chromosomes as the earliest cytogenetic events. In contrast, the gains and losses of whole chromosomes in group 4 occurred through multiple simultaneous events leading to a near-triploid chromosome number.

CONCLUSION

The finding of different temporal patterns for the acquisition of genomic imbalances in high-risk and low-risk NBs lends strong support to the hypothesis that these tumours are biologically diverse entities, evolving through distinct genetic mechanisms.

SKY analysis of childhood neural tumors and cell lines demonstrates a susceptibility of aberrant chromosomes to further rearrangements

Malignant solid tumors are commonly characterized by a large number of complex structural and numerical chromosomal alterations, which often reflect the level of genomic instability and can be associated with disease progression. The aim of this study was to evaluate whether chromosomes that harbor primary aberrations have a higher susceptibility to accumulate further alterations. We used spectral karyotyping (SKY), to compare the individual chromosomal instability of two chromosome types: chromosomes that have a primary aberration and chromosomes without an aberration, in 13 primary childhood neural tumors and seven cell lines. We found that chromosomes that contain a primary aberration are significantly (p-value<0.001) more likely to gain further structural rearrangements or to undergo numerical changes (22.6%, 36 of 159 chromosomes) than chromosomes with no initial aberration (4.9%, 54 of 1099 chromosomes). These results are highly suggestive that aberrant chromosomes in solid tumors have a higher susceptibility to accumulate further rearrangements than "normal" chromosomes.

Markers improve clustering of CGH data

MOTIVATION

We consider the problem of clustering a population of Comparative Genomic Hybridization (CGH) data samples using similarity based clustering methods. A key requirement for clustering is to avoid using the noisy aberrations in the CGH samples.

RESULTS

We develop a dynamic programming algorithm to identify a small set of important genomic intervals called markers. The advantage of using these markers is that the potentially noisy genomic intervals are excluded during the clustering process. We also develop two clustering strategies using these markers. The first one, prototype-based approach, maximizes the support for the markers. The second one, similarity-based approach, develops a new similarity measure called RSim and refines clusters with the aim of maximizing the RSim measure between the samples in the same cluster. Our results demonstrate that the markers we found represent the aberration patterns of cancer types well and they improve the quality of clustering significantly.

AVAILABILITY

All software developed in this paper and all the datasets used are available from the authors upon request.

Analysis of genetic abnormalities provides insights into genetic evolution of hyperdiploid myeloma

Aneuploidy is ubiquitous in human cancer and is seen as whole chromosome gains and losses, unbalanced translocations and inversions, duplications, deletions and loss of heterozygosity. Within this complexity, some subgroups of aneuploid tumors emerge as distinct biological and clinical entities. Hyperdiploid myeloma (H-MM), characterized by hyperdiploid chromosome numbers because of nonrandom trisomies, is one such example. We undertook a comprehensive survey of the karyotypes of a large number of H-MM (n = 469) to describe fully genomic instability in these tumors, to dissect pathways of genetic evolution, and identify distinct subgroups based on their genetic changes. While selective pressure apparently favors the emergence of clones with gains of chromosomes 3, 5, 7, 9, 11, 15, 19, and 21, a background of ongoing genomic instability results in gains of other chromosomes, albeit at a much lower prevalence. A deduced temporal analysis of these karyotypes indicates that selected gains are early events. Other events occurring later in the course of the disease include secondary chromosome translocations and monosomies. The development of these genetic aberrations is thus highly ordered and undoubtedly of biological relevance. Within this framework, we propose a model of genetic evolution in H-MM.

Clonal and non-clonal chromosome aberrations and genome variation and aberration

The theoretical view that genome aberrations rather than gene mutations cause a majority of cancers has gained increasing support from recent experimental data. Genetic aberration at the chromosome level is a key aspect of genome aberration and the systematic definition of chromosomal aberrations with their impact on genome variation and cancer genome evolution is of great importance. However, traditionally, efforts have focused on recurrent clonal chromosome aberrations (CCAs). The significance of stochastic non-clonal chromosome aberrations (NCCAs) is discussed in this paper with emphasis on the simple types of NCCAs that have until recently been considered "non-significant background". Comparison of various subtypes of transitional and late-stage CCAs with simple and complex types of NCCAs has uncovered a dynamic relationship among NCCAs, CCAs, overall genomic instability, and karyotypic evolution, as well as the stochastic nature of cancer evolution. Here, we review concepts and methodologies to measure NCCAs and discuss the possible causative mechanism and consequences of NCCAs. This study raises challenging questions regarding the concept of cancer evolution driven by stochastic chromosomal aberration mediated genome irregularities that could have repercussions reaching far beyond cancer and organismal genomes.

New probabilistic network models and algorithms for oncogenesis

Chromosomal aberrations in solid tumors appear in complex patterns. It is important to understand how these patterns develop, the dynamics of the process, the temporal or even causal order between aberrations, and the involved pathways. Here we present network models for chromosomal aberrations and algorithms for training models based on observed data. Our models are generative probabilistic models that can be used to study dynamical aspects of chromosomal evolution in cancer cells. They are well suited for a graphical representation that conveys the pathways found in a dataset. By allowing only pairwise dependencies and partition aberrations into modules, in which all aberrations are restricted to have the same dependencies, we reduce the number of parameters so that datasets sizes relevant to cancer applications can be handled. We apply our framework to a dataset of colorectal cancer tumor karyotypes. The obtained model explains the data significantly better than a model where independence between the aberrations is assumed. In fact, the obtained model performs very well with respect to several measures of goodness of fit and is, with respect to repetition of the training, more or less unique.

Distance-based clustering of CGH data

MOTIVATION

We consider the problem of clustering a population of Comparative Genomic Hybridization (CGH) data samples. The goal is to develop a systematic way of placing patients with similar CGH imbalance profiles into the same cluster. Our expectation is that patients with the same cancer types will generally belong to the same cluster as their underlying CGH profiles will be similar.

RESULTS

We focus on distance-based clustering strategies. We do this in two steps. (1) Distances of all pairs of CGH samples are computed. (2) CGH samples are clustered based on this distance. We develop three pairwise distance/similarity measures, namely raw, cosine and sim. Raw measure disregards correlation between contiguous genomic intervals. It compares the aberrations in each genomic interval separately. The remaining measures assume that consecutive genomic intervals may be correlated. Cosine maps pairs of CGH samples into vectors in a high-dimensional space and measures the angle between them. Sim measures the number of independent common aberrations. We test our distance/similarity measures on three well known clustering algorithms, bottom-up, top-down and k-means with and without centroid shrinking. Our results show that sim consistently performs better than the remaining measures. This indicates that the correlation of neighboring genomic intervals should be considered in the structural analysis of CGH datasets. The combination of sim with top-down clustering emerged as the best approach.

AVAILABILITY

All software developed in this article and all the datasets are available from the authors upon request.

CONTACT

juliu@cise.ufl.edu.

The order of genetic events associated with colorectal cancer progression inferred from meta-analysis of copy number changes

To identify chromosomal aberrations that differentiate among the Dukes' stages of colorectal cancer (CRC) as well as those that are responsible for the progression into liver metastases, we performed a meta-analysis of data obtained from 31 comparative genomic hybridization (CGH) studies comprising a total of 859 CRCs. Individual copy number profiles for 373 primary tumors and 102 liver metastases were recorded and several statistical analyses, such as frequency, multivariate logistic regression, and trend tests, were performed. In addition, time of occurrence analysis was applied for the first time to copy number changes identified by CGH, and each genomic imbalance was thereby classified as an early or late event in colorectal tumorigenesis. By combining data from the different statistical tests, we present a novel genetic pathway for CRC progression that distinguishes the Dukes' stages and identifies early and late events in both primary carcinomas and liver metastases. Results from the combined analyses suggest that losses at 17p and 18 and gains of 8q, 13q, and 20 occur early in the establishment of primary CRCs, whereas loss of 4p is associated with the transition from Dukes' A to B-D. Deletion of 8p and gains of 7p and 17q are correlated with the transition from primary tumor to liver metastasis, whereas losses of 14q and gains of 1q, 11, 12p, and 19 are late events. We supplement these findings with a list of potential target genes for the specific alterations from a publicly available microarray expression dataset of CRC.

Drugs aimed at targeting characteristic karyotypic phenotypes of cancer cells

The karyotypic features of cancer cells have not been a particular focus of anticancer drug targeting either as guidance for treatment or as specific drug targets themselves. Cancer cell lines typically have considerable, characteristic, and variable chromosomal aberrations. Here, we consider small-molecule screening data across the National Cancer Institute's 60 tumor cell line drug screening panel (NCI-60) analyzed for specific association with karyotypic variables (numerical and structural complexity and heterogeneity) determined for these same cell lines. This analysis is carried out with the aid of a self-organizing map allowing for a simultaneous assessment of all screened compounds, revealing an association between karyotypic variables and a unique part of the cytotoxic response space. Thirteen groups of compounds based on related specific chemical structural motifs are identified as possible leads for anticancer drug discovery. These compounds form distinct groups of molecules associated with relatively unexplored regions of the NCI-60 self-organizing map where anticancer agents currently standard in the clinic are not present. We suggest that compounds identified in this study may represent new classes of potential anticancer agents.

Transposable elements – Is there a link between evolution and cancer?

Currently, the most predominant theory concerning the formation of cancer is that it is a genetic accident. Accordingly, various agents are thought to cause DNA damage which then subsequently activates oncogenes and inactivates tumor suppressor genes. This article, however, describes a theory that interprets cancer as a misguided adaptation. Stressors, which cannot be compensated for with the usual cell possibilities might arouse evolutionary mechanisms intended to create new protein variants. One of these is the activation of transposable elements which leads to a reformatting of the genome. The result of this process is either a cell that survives very well under stress (and will, therefore, never be detected), a dead cell (in case the process is ineffective), or a more or less abnormal and harmful cell that builds up a new but cancerous organ. This theory explains the complex genetic alterations which are present in almost all cancer cells. It also explains the action of non-mutagenic carcinogens. As part of the reformatting process of the cancer cell genome, activation of oncogenes and inactivation of tumor suppressor genes are not stochastic events but the result of an unlucky genomic composition.

Molecular subtypes of bladder cancer: Jekyll and Hyde or chalk and cheese?

Cancer of the bladder shows divergent clinical behaviour following diagnosis and it has been proposed that two major groups of tumours exist that develop via different molecular pathways. Low-grade, non-invasive papillary tumours recur frequently, but patients with these tumours do not often suffer progression of disease to muscle invasion. In contrast, tumours that are invading muscle at diagnosis are aggressive and associated with significant mortality. Molecular studies have identified distinct genetic, epigenetic and expression changes in these groups. However, it is not yet clear whether there is direct progression of low-grade superficial tumours to become invasive (a Jeckell and Hyde scenario) or whether in those patients who apparently progress from one form of the disease to the other, different tumour clones are involved and that the two tumour groups are mutually exclusive ('chalk and cheese'). If the latter is true, then attempts to identify molecular markers to predict progression of low-grade superficial bladder tumours may be fruitless. Similarly, it is not clear whether other subgroups of tumours exist that arise via different molecular pathways. There is now a large amount of molecular information about bladder cancer that facilitates examination of these possibilities. Some recent studies provide evidence for the existence of at least one further group of tumours, high-grade superficial papillary tumours, which may develop via a distinct molecular pathway. Patients with such tumours do show increased risk of disease progression and for these there may exist a real progression continuum from non-invasive to invasive. If this is the case, definition of the molecular signature of this pathway and improved understanding of the biological consequences of the events involved will be pivotal in disease management.

Statistical dissection of genetic pathways involved in prostate carcinogenesis

Molecular markers that could stratify prostate cancer patients according to risk of disease progression would allow a significant improvement in the management of this clinically heterogeneous disease. In the present study, we analyzed the genetic profile of a consecutive series of 51 clinically confined prostate carcinomas and 27 benign prostatic hyperplasias using comparative genomic hybridization (CGH). We then added our findings to the existing literature data in order to perform a meta-analysis on a total of 294 prostate cancers with detailed CGH and clinicopathological information, using multivariate statistical methods that included principal component, hierarchical clustering, time of occurrence, and regression analyses. Whereas several genomic imbalances were shared by organ-confined, locally invasive, and metastatic prostate cancers, 6q and 10q losses and 7q and 8q gains were significantly more frequent in patients with extra-prostatic disease. Regression analysis indicated that 8q gain and 13q loss were the best predictors of locally invasive disease, whereas 8q gain and 6q and 10q losses were associated with metastatic disease. We propose a genetic pathway of prostate carcinogenesis with two distinct initiating events, namely, 8p and 13q losses. These primary imbalances are then preferentially followed by 8q gain and 6q, 16q, and 18q losses, which in turn are followed by a set of late events that make recurrent and metastatic prostate cancers genetically more complex. We conclude that significant differences exist in the genetic profile of organ-confined, locally invasive, and advanced prostate cancer and that genetic features may carry prognostic information independently of Gleason grade.

Ewing tumours and synovial sarcomas have critical features of karyotype evolution in common with epithelial tumours

We have analysed the accumulated cytogenetic data on karyotypic evolution in Ewing tumours (ET) and synovial sarcomas (SS). Both tumour types frequently show balanced translocations, t(11;22) and t(X;18), respectively, that result in specific fusion genes. The analyses revealed +8, +12, +1q, and 16q- as important secondary changes to t(11;22) in ET and the imbalances showed a distinct temporal order. By principal component analysis, one major karyotypic pathway dominated by gains and one minor dominated by losses were identified. The kartyotypic evolution pattern in SS was less distinct. Both ET and SS showed a power law distribution of the number of acquired aberrations, which in both tumour types conformed to a distribution with an exponent equal to 1. Similar distributions are frequently found in epithelial tumours. ET and SS differ in this respect from other malignancies with balanced translocations resulting in fusion genes, which typically show a power law distribution of the number of acquired aberrations with exponents close to 2. This suggests that chromosome changes in ET and SS may develop through mechanisms more similar to those in epithelial tumours lacking recurrent balanced rearrangements than in haematological malignancies characterised by balanced translocations leading to fusion genes.