INTRODUCTION: One important characteristic of solid tumors is heterogeneity at multiple levels of genetic and non-genetic organization. This can include gene mutations, epigenetic alterations, copy number changes, and chromosomal aberrations. Collectively, these alterations contribute as parts of a genome-defined system. Thus, when genetic information is passed from mother to daughter cell in the context of cancer evolution, in contrast to normal cellular processes, an altered system inheritance is often transmitted. When the genome of a somatic cell is highly unstable, such as during certain phases of cancer initiation and progression, many novel alterations to the genome can be introduced in a short timeframe, effectively resulting in the macro-evolution of the somatic cell population (i.e., through the transition stages of cancer, including transformation, metastasis, and drug resistance). Unfortunately, these continually introduced, non-clonal alterations to the cell’s genetic information have often been described as background “noise” that does not function significantly in cancer. Rather, the driving force of cancer has largely been attributed to the accumulation of gene mutations in several key, driver genes. Despite the presumed significance of these driver genes by the gene mutation and clonal evolutionary theories of cancer, recent sequencing efforts have failed to identify common driver genes in the majority of cancer types. Based on this fact, and on the overwhelming presence of non-clonal alterations at multiple levels of organization in the cells comprising tumors, the paradigm of cancer research requires re-examination. A better understanding of genome-level heterogeneity is necessary, as the genome, rather than individual genes, defines system boundaries and unifies the diverse individual molecular mechanisms of cancer through their contribution to major evolutionary transitions. Because inheritance is traditionally defined as a precise process of relaying bio-information with extreme low frequencies of errors, it is challenging to explain how genetics work in cancer evolution. It is thus timely to consider that potentially novel processes of inheritance occur in many types of cancer. The maintenance of a massive extent of multi-level heterogeneity in the cells of solid tumors over generations suggests that a less precise process is taking place. We have described this with a new term, “fuzzy inheritance,” wherein a range of variants, rather than specific variants (such as specific gene mutations or chromosomal aberrations), is recapitulated in the cell division process. This study aimed to elucidate the mechanism of fuzzy inheritance by examining the relationship between genome instability-linked karyotypic heterogeneity and growth heterogeneity, based on single-cell analysis of an in vitro cell culture model. By demonstrating that increased genome-level heterogeneity is reflected by increased and more variable levels of growth heterogeneity, it was hoped to establish that fuzzy inheritance correctly explains the maintenance of high levels of heterogeneity in these somatic cell populations. An example of this phenomenon was also studied in giant cancer cells, as they undergo division processes which appear to contribute to and facilitate genome instability. METHODS: To examine these concepts, various cellular profiling methods were used, including in-situ cell growth, cellular morphological comparison, and karyotype analysis. We first quantified the extent of variation in the growth rates of single cells; by selecting the fastest- and slowest-growing colonies from the parent population, and examining the extent to which growth heterogeneity was passed in subsequent generations of cells, the correlation between genome-level heterogeneity (as reflected by the karyotype) and growth heterogeneity was determined. We then examined an extreme example of fuzzy inheritance, wherein giant cancer cells containing massive amounts of DNA undergo extremely abnormal cell division events, yielding many normal-sized daughter cells with genomes significantly different from those of both the parent cell and other daughter cells. By studying the frequency and other aspects of these cells in two unequally stable cell lines, we sought to gain insight on one specific mechanism of fuzzy inheritance. RESULTS: The data suggested that fuzzy inheritance can be demonstrated in multiple cell culture models. The extent and variability of karyotypic heterogeneity was reflected by those of growth heterogeneity, indicating the karyotype’s importance in facilitating cancer evolutionary processes. Moreover, the cells with giant nuclei can generate diverse genome-level heterogeneity. DISCUSSION: Because fuzzy inheritance allows for the less precise passage of bio-information over generations in cancer cell populations, and for the effective introduction of numerous alterations to the genome in often brief spans of time, the cell population can constantly increase its evolutionary potential, which is essential for the major transition steps of cancer evolution. The mechanism of fuzzy inheritance should be explored further, due to its clear importance in the processes underlying cancer initiation, progression, and drug resistance.