Modeling and targeting general and chromosome-specific aneuploidy in cancer

In this review, Goldberg et al. describe the contextual dependencies of aneuploidy and its contribution to cell-intrinsic and cell-extrinsic phenotypes in cancer. They also discuss recent advances in detecting, modeling, and targeting aneuploidy, reflecting the field's efforts to leverage aneuploidy-induced vulnerabilities into effective strategies in cancer diagnosis and therapy.

Abstract

Throughout the last century, aneuploidy has been cemented as a hallmark of cancer. Although the association of aneuploidy with tumorigenesis has been well established, the role of these genetic imbalances in tumor formation has only recently begun to be elucidated. Advancements in genomics have revealed the complexity and context dependence of the effect of aneuploidy on cancer growth, while developments in genetic editing have allowed for proper modeling of specific aneuploidies. In this review, we discuss the key factors to consider when studying the role of aneuploidy in cancer and the tools that are available to do so. We then highlight recent studies that establish phenotypic contributions of aneuploidy to tumorigenicity. In particular, we highlight how general aneuploidy and chromosomal instability affect the tumor microenvironment and how specific chromosomal alterations, including the loss of chromosome 9p and the gain of chromosomes 8q and 1q, influence tumor behavior and therapeutic responses. Finally, we emphasize the potential of targeting aneuploidy-induced vulnerabilities to improve cancer treatment outcomes.

General and chromosome-specific aneuploidy

In the cancer field, aneuploidy is considered to be an abnormal number of whole chromosomes or chromosome arms. The relationship between aneuploidy and tumorigenesis was first established by David von Hansemann (1890) and Theodor Boveri (2008) >100 years ago, and technological advances since then have enabled further insights into the phenomenon (Hardy and Zacharias 2005). Sequencing of large numbers of tumors has shown that ∼90% of tumors exhibit aneuploidy (Taylor et al. 2018). Importantly, studies using cytogenetics, comparative genomic hybridization, and next-generation sequencing all noticed the existence of specific karyotypic patterns that are recurrent across tumors (Bown 1992; Zitzelsberger et al. 1997; Taylor et al. 2018). For example, chromosome arm 17p is lost in 35% of tumors, whereas 8q is gained in 33% of tumors. These chromosome-specific patterns of aneuploidy are also tumor type-specific, with certain chromosomes being frequently gained in one cancer type while commonly being lost in another. This observation has affected how aneuploidy in cancer is studied. Therefore, in this review, we separately focus on two distinct aneuploidy types: general aneuploidy, referring to a karyotype that deviates from a multiple of the haploid state (23 in humans), and chromosome-specific aneuploidy, defined as the gain or loss of a specific chromosome or chromosome arm.

More specifically, on the one hand, here we refer to general aneuploidy in cancer as the presence of an abnormal number of chromosomes regardless of which specific chromosomes are gained or lost. On the other hand, we refer to chromosome-specific aneuploidy (CSA), where we specifically highlight chromosomes that tend to be most frequently gained or lost in particular tumor types, such as chromosome 9p loss or 1q gain. These two types of aneuploidy have different effects, consequences, and associated phenotypes in tumors. General aneuploidy leads to changes in cellular homeostasis and increased cellular stress that are independent of the identity of the altered chromosomes. In contrast, chromosome-specific aneuploidy produces consequences that are directly tied to the function and content of the affected chromosome. In a typical solid tumor karyotype, both types of aneuploidy coexist: In a highly aneuploid karyotype, usually containing >46 chromosomes, there are numerous gains and losses (general aneuploidy), among which certain specific alterations are consistently observed across patients (chromosome-specific aneuploidy) (Fig. 1A).

Figure 1.

(A) In the cancer field and in this review, aneuploidy is considered to be an alteration in the copy number of whole chromosomes or chromosome arms. Focal alterations—that is, gains or losses that encompass a genomic region smaller than a chromosome arm—are also included in copy number alteration studies and often are referred to as segmental aneuploidy. (B) General aneuploidy in human cancer. Frequency distributions of the total number of chromosome arm gains and losses in pan-cancer and selected tumor types from The Cancer Genome Atlas (TCGA) (data are from Taylor et al. 2018). (C) Chromosome-specific aneuploidy (CSA). Pan-cancer and tumor type-specific patterns of arm-level gains and losses from TCGA.

Although not the primary focus of this review, some of the concepts described in the context of CSA apply also to focal events; that is, gains or losses of segments of genomes smaller than chromosome arms.

It is important to note that chromosomal instability (CIN), defined as an increased rate of chromosome missegregation, is tightly related to aneuploidy, with each being capable of inducing and exacerbating the other (Thompson et al. 2010; Potapova et al. 2013). Consequently, many phenotypes observed in the context of CIN or aneuploidy may in fact be due to either process—or a combination of both. In this review, we include findings related to CIN within the discussion of general aneuploidy but distinguish between the two when possible. In general, due to its mechanistic nature and ability to be induced, CIN is often induced in experimental studies and interpreted as a causal driver of phenotypic outcomes. In contrast, general aneuploidy is typically treated as a correlative feature observed in tumor data sets and associated with various clinical outcomes and cellular characteristics in observational studies.

As aneuploidy became recognized as a hallmark of cancer, interest has been directed toward its prognostic value as well. Flow cytometry studies of tumor samples and more recent pan-cancer sequencing studies have shown that general aneuploidy correlates with metastasis and poor survival outcomes in diverse malignancies, including breast, prostate, colon, and lung cancers and melanoma (Merkel and McGuire 1990; Danielsen et al. 2016; Hieronymus et al. 2018; Jeon et al. 2024). Metrics such as the global copy number alteration burden or “aneuploidy score” are widely used to quantify the degree of aneuploidy in tumor genomes and often outperform traditional prognostic factors like stage, grade, or point mutations in explaining survival differences. For example, in lung cancer, aneuploidy predicts poor patient survival independent of age, smoking, histologic subtype, adjuvant therapy, and tumor stage (Jamal-Hanjani et al. 2017). Additionally, certain individual or combined CSAs have been found to be prognostic (Kou et al. 2020; Shukla et al. 2020; Vasudevan et al. 2020). Just as CSA frequencies vary between cancers, different CSAs have been proposed as prognostic biomarkers depending on tumor type. For example, 8q gain can predict poor survival in prostate and kidney cancers, whereas loss of 3p can predict poor prognosis in head and neck cancers (Ribeiro et al. 2006; Mehrazin et al. 2018; Shaikh et al. 2020). Interestingly, individuals with Down syndrome, caused by trisomy 21, are more likely to develop leukemias but less likely to develop many solid tumors than individuals without Down syndrome (Moretti et al. 2025).

The widespread frequency and prognostic value of aneuploidy point to a significant correlative relationship between an abnormal karyotype and tumorigenesis but introduce the greater question of whether aneuploidy is a cause or effect of transformation. The gain or loss of an entire chromosome alters the dosage of thousands of genes at a time, and the levels of the encoded transcripts and proteins generally mirror these dosage changes (Stingele et al. 2012). The genomic and proteomic imbalances caused by aneuploidy result in proteotoxic, DNA replication, and metabolic stresses, leading to a decrease in cell fitness under normal conditions. However, in the context of the highly mutational and selective process of tumorigenesis, cells are able to adapt to tolerate aneuploidy and even benefit from it (Stingele et al. 2012; Zhu et al. 2018; Chunduri and Storchová 2019; Zhou et al. 2020). It is believed that the intratumoral heterogeneity introduced by CIN is subject to both positive and negative selection at different points of tumorigenesis to generate the recurrent CSA patterns observed across tumor types and stages (Sansregret and Swanton 2017; Klockner and Campbell 2024). CSA affects cell function by altering the copy number of genes located on the gained or lost chromosome. Accordingly, CSA is thought to promote tumorigenesis by altering the dosage of oncogenes and tumor suppressors (Davoli et al. 2013) and to contribute to metastasis, immune evasion, and drug resistance by impacting relevant genes (Shukla et al. 2020; Zhakula-Kostadinova and Taylor 2024).

Although paradoxical with its occurrence in a disease of uncontrolled cell growth, general aneuploidy has been found to negatively affect cell fitness. The stresses induced by aneuploidy confer cells with weaknesses that can be therapeutically targeted. Additionally, CSAs drive tumorigenesis by altering the dosage of specific genes to activate certain tumor-specific oncogenic pathways that can be inhibited to selectively kill cancer cells and prevent cancer progression. In the following sections, we discuss factors that should be considered in order to identify the contexts in which aneuploidy is likely to contribute to cancer growth and techniques for engineering aneuploidy. We then highlight recently discovered phenotypic contributions of aneuploidy to tumorigenesis and summarize the therapeutic opportunities currently being investigated.

Contextual considerations for the study of aneuploidy in cancer

Studies of the causes and consequences of aneuploidy have yielded confounding results, suggesting that the effect of aneuploidy on cells is not universal and revealing that uncovering why it is a hallmark of cancer requires the consideration of many variables. Both general aneuploidy and CSA have been found to either promote or suppress malignancy depending on the cellular or genetic context in which they occur or the phenotype being observed. The results of previous studies as well as tumor and cell line data sets can provide insight for selecting the context in which to study aneuploidy in order to reveal its tumorigenic role.

Contextual dependency of general aneuploidy

As mentioned previously, CIN induces general aneuploidy that initially reduces cellular fitness under normal conditions but generates karyotypic heterogeneity from which advantageous clones can be selected to promote tumor growth. In order to study the role of aneuploidy in cancer, it would be ideal to induce aneuploidy in a way that is representative of what occurs in human disease. Challenging this ability is the fact that the origin of CIN in tumors is still being elucidated. Early studies of chromosome missegregation-inducing mechanisms involved interfering with the function of genes involved in sister chromatid cohesion, centrosome replication, kinetochore–microtubule attachment, and the spindle assembly checkpoint (SAC). Interestingly, both the reduction and overexpression of SAC genes can lead to aneuploidy-induced tumorigenesis in mice by weakening or hyperactivating the SAC, respectively (Holland and Cleveland 2009; Schvartzman et al. 2010; Vasudevan et al. 2021; Hosea et al. 2024). Similarly, both reduction and overexpression of genes involved in chromosome segregation are found in human tumors (Simonetti et al. 2019). Germline mutations in key SAC regulatory genes are associated with increased tumor incidence in individuals with mosaic variegated aneuploidy (Malumbres and Villarroya-Beltri 2024), and rare mutations in BUB1 have been observed in pancreatic cancer and early cases of colorectal cancer (Hernando et al. 2001; Hempen et al. 2003; Taylor et al. 2018). On the other hand, the overexpression of genes such as MAD2L1, BUB1, and PLK1 has been observed in colorectal, lung, and breast cancers (Yuan et al. 2006; He et al. 2019; Ding et al. 2020). Several common oncogenes and tumor suppressors, including MYC (Littler et al. 2019; Rohrberg et al. 2020), APC (Rusan and Peifer 2008), RB1 and KRAS (Coschi and Dick 2012), have been implicated in causing CIN by either hyperactivating proliferation or playing more direct roles in chromosome segregation.

Although it is clear that CIN is a characteristic of many tumors, the relationship between CIN level and malignancy is not linear. Perturbations in some CIN-inducing genes are weakly able to induce tumorigenesis, and perturbations in other CIN-inducing genes only promote tumor growth in the context of other oncogenic mutations or after treatment with carcinogens (Holland and Cleveland 2009; Schvartzman et al. 2010; Vasudevan et al. 2021; Hosea et al. 2024). These findings indicate that low levels of CIN alone are insufficient for tumorigenesis. At the same time, extreme levels of CIN are toxic, as irreparable abnormalities induce senescence or apoptosis, and the resultant aneuploidy-induced stresses decrease fitness. These observations have led to the generation of the “just right” hypothesis, which proposes that only a moderate level of CIN provides the proper level of genetic instability and variability for tumorigenesis without causing cell death (Sansregret and Swanton 2017; Chunduri and Storchová 2019; Vasudevan et al. 2021; Hosea et al. 2024). Examples of this nonlinear relationship between CIN and tumorigenic fitness also exist in human tumors; for example, in gastroesophageal adenocarcinoma, intermediate levels of aneuploidy predict worse prognosis compared with either low or high levels, suggesting that tumors with moderate genomic instability may achieve a balance between sufficient adaptability to thrive and diminished vulnerability to cell stress (Davison et al. 2014).

The genetic background of a cell is likely to influence the effect that aneuploidy has on its physiology and ultimate tumorigenicity. The toxicity of CIN and aneuploidy may only allow for the survival of cells that contain additional alterations that make these traits more tolerable. The best-studied gene whose abrogation allows aneuploid cells to propagate is TP53, which is thought to cause cell cycle arrest, senescence, and death mainly in response to the DNA damage caused by aneuploidy (Marques and Kops 2023). However, even the activation of p53 in response to aneuploidy itself is context-dependent, varying between organisms, culture conditions, and how the aneuploidy arises. Thus, it is possible that the effect of aneuploidy on cell division and viability may differ between p53 wild-type and mutant models. The identification of additional mutations that could alleviate aneuploidy-induced stresses is an active area of research, with multiple studies pointing to a role for MAPK signaling in increasing the proliferation of aneuploid cells (Hintzen et al. 2024; Zerbib et al. 2024). Finally, tumors with microsatellite instability or DNA polymerase ε mutations exhibit a negative correlation between aneuploidy and mutation rate, indicating that cancers may only be able to tolerate and take advantage of one mode of genomic instability at a time (Taylor et al. 2018).

The karyotypic context of aneuploidy can also influence the effect it has on a cell. Whether or not a tumor has experienced whole-genome doubling (WGD) is reflected in the degree and types of aneuploidy observed (Dewhurst et al. 2014; Kuznetsova et al. 2015); WGD⁺ tumors and cell lines display a larger range of aneuploidy, more whole-chromosome aneuploidy, and more losses compared with WGD⁻ samples, indicating that WGD⁺ cells have a higher tolerance for genetic imbalances (Prasad et al. 2022). Finally, when studying the effects of general aneuploidy on cancer, it is important to note that chromosomal gains and losses can have different effects on cellular physiology, with losses tending to be more harmful to cell viability, proliferation, and genomic stability (Chunduri et al. 2022; Hintzen et al. 2022; Torres 2023).

The effect of aneuploidy can depend on the cell type in which it is induced, with aneuploidy generally decreasing the malignancy of primary and transformed cells while increasing the oncogenic potential of pluripotent stem cells (Williams et al. 2008; Zhang et al. 2016; Sheltzer et al. 2017; Vasudevan et al. 2020; Xiao et al. 2024). Additionally, mouse models of CIN appear to cause tumorigenesis in various tissues depending on the genetic perturbation introduced (Holland and Cleveland 2009; Schvartzman et al. 2010; Vasudevan et al. 2021; Hosea et al. 2024). This raises the question of whether CIN and general aneuploidy contribute to tumorigenesis in a tissue-specific manner, just as deregulated oncogenic pathways vary across tumor types. One context where this might occur is in the intestines of mice. Two individual studies found that although APC^Min/+ mice form multiple adenomas in the small intestine but few to none in the colon, increasing CIN in APC^Min/+ mice by introducing BubR1 or Mps1 mutations decreased adenoma formation in the small intestine while increasing it in the colon. These findings suggest that the small intestine is more sensitive to both the tumor-promoting and tumor-suppressing effects of CIN than the large intestine (Rao et al. 2005; Hoevenaar et al. 2020). This is further supported by the observation that different tumor types have different frequency distributions of gains and losses (Fig. 1B). Studies introducing the same levels of aneuploidy and CIN across various tissues and cell types are necessary to establish any tissue-specific effect of general aneuploidy on tumorigenesis.

Finally, whether general aneuploidy and CIN appear to be cancer-promoting or cancer-suppressing depends on the phenotype being observed. Due to the many stresses it causes, aneuploidy decreases the proliferation rate of mammalian cells under standard conditions in vitro (Williams et al. 2008; Stingele et al. 2012). However, certain trisomies were found to increase the aggressiveness of human cells under stressful conditions such as hypoxia and serum starvation (Rutledge et al. 2016). How CIN and aneuploidy affect metastasis and immune response are still open questions. Broadly, CIN promotes metastasis, whereas the effects of aneuploidy are more chromosome-specific (Vasudevan et al. 2020; Zheng et al. 2023). CIN and aneuploidy seem to activate innate immune response pathways, yet clinical samples show high aneuploidy correlating with low immune cell infiltration, leading to the current hypothesis that aneuploid tumors become immune-evasive during tumor progression through adaptation and selection (Bakhoum and Cantley 2018; Kuang and Li 2022). Because these phenotypes depend on the tumor microenvironment in addition to cell-intrinsic characteristics, the development of immune-competent mouse models of aneuploid yet chromosomally stable tumors will be important for differentiating the effects of CIN and aneuploidy on these processes. Finally, CIN increases drug resistance by increasing karyotypic heterogeneity (Kwong et al. 2017; Ippolito et al. 2021; Lukow and Sheltzer 2022), though the aneuploidy-induced stresses discussed above may confer cells with specific therapeutic weaknesses (see “Therapeutic Strategies for Targeting Aneuploidy in Cancer” below). In summary, the effect of general aneuploidy on malignancy depends on how the aneuploidy arises, the presence of additional oncogenic and aneuploidy-tolerating mutations, the cell type in which aneuploidy is introduced, and the phenotype being observed, such as cell proliferation, tumor initiation, tumor progression, or cancer immune evasion.

Contextual dependency of chromosome-specific aneuploidy

The effects of CSA on tumorigenesis vary depending on the same contexts as general aneuploidy. Proliferative drivers are tissue-specific; for example, KRAS is mutated in lung, pancreatic, and colorectal cancers but is rarely mutated in breast or renal cancer (Haigis et al. 2019). In a similar way, CSA patterns that alter the dosage of such drivers are also tissue-specific (Sack et al. 2018). With regard to genetic context, CSAs can interact with mutations, fusions, and other CSAs. Last, the specific genes whose dosages are affected by chromosomal gains or losses all have different functions, leading to these changes contributing to various cancer phenotypes. Publicly available patient and cell line data sets can be used to identify likely driver aneuploidies and their contexts.

The observation that each chromosome is gained and lost at different frequencies in different tumor types suggests that the effect of each CSA on tumor growth depends on the cell type in which it occurs (Fig. 1C; Taylor et al. 2018). Both positive and negative selection act on the random aneuploidies introduced by CIN to form the CSA patterns observed across tumor types (Jubran et al. 2024; Klockner and Campbell 2024). One model that explains these tissue-specific patterns proposes that CSA promotes tumorigenesis by altering the copy number of tissue-relevant proliferative drivers and suppressors. For example, the predisposition of Down syndrome individuals to leukemia is hypothesized to be due to the amplification of chromosome 21-encoded genes leading to the disruption of fetal hematopoiesis (Baruchel et al. 2023), whereas the protection from solid tumors has been attributed to trisomic genes limiting tumor angiogenesis (Zorick et al. 2001; Sussan et al. 2008; Shin et al. 2014). Further supporting this hypothesis is the finding that proliferative drivers and suppressors are tissue-specific, and their distribution across the genome can predict the CSA patterns observed across tumor types (Davoli et al. 2013; Sack et al. 2018). Notably, the specificity of CSA patterns is observed even among cells that share tissue origin or developmental lineage, as molecular subtypes of most tumors exhibit similar CSA patterns. One exception is ovarian cancer, where high-grade serous carcinoma, clear cell carcinoma, and endometrioid carcinoma differ in both their overall aneuploidy levels and the CSA patterns that they display (Tapper et al. 1997; Uehara et al. 2015). Finally, the CSA patterns in tumors appear to mimic and amplify the gene expression patterns of the normal tissues from which they arise; chromosomal deletions tend to occur on chromosomes with fewer actively expressed genes, whereas amplifications are more common on chromosomes rich in actively expressed genes (Patkar et al. 2021). This could serve to intensify the expression of tissue-specific proliferative drivers and suppressors. Alternatively, this pattern could reflect that amplifying pre-existing expression programs is more tolerable than increasing the dosage of silent genes or decreasing that of active ones.

The context dependency of the role of CSA in tumorigenesis can be further specified by the status of other genetic features. Analysis of large data sets can allow for the determination of co-occurrence or mutual exclusivity of a CSA with a certain genomic alteration. The most intuitive reason for a CSA to co-occur with a mutation would be to amplify the mutation's effect by either increasing the copy number of an activated oncogene or decreasing the copy number of a functional tumor suppressor. However, there is currently limited evidence of aneuploidy serving as a mechanism of allelic imbalance (Bielski and Taylor 2021; Besedina and Supek 2024). Nevertheless, CSA patterns can differ between tumors driven by distinct oncogenic mutations, suggesting that specific mutations may influence the selection of particular chromosomal alterations (Gao et al. 2021). Additionally, CSAs can interact with translocations. For example, chromosome 8 gain commonly occurs in Ewing sarcoma driven by the EWS-FLI1 fusion oncogene. This correlation has been mechanistically explained by chromosome 8 gain increasing the dosage of RAD21, which relieves EWS-FLI1-induced replication stress (Su et al. 2021). Finally, the frequency of a CSA can be tied to the presence of another CSA, such as the co-occurrence of chromosome 7 gain and 10p loss in glioblastoma (Nair et al. 2024). The frequency of such genetic interactions can differ depending on WGD status; some CSAs that co-occur in WGD⁺ tumors are mutually exclusive in WGD⁻ tumors and vice versa (Prasad et al. 2022). Due to the vast complexity of the human genome, pinpointing which alterations drive the effects of CSA remains a significant challenge. One approach is to compare CSA patterns across tumors grouped by established molecular subtypes or the presence of clinically relevant mutations. Additionally, tumor evolutionary analyses can be used to determine the sequence in which mutation and copy number events occur to provide insights into the genetic background required for a CSA to be beneficial (Gerstung et al. 2020; Watkins et al. 2020).

A CSA can appear to be beneficial or not depending on the phenotype being observed. For example, HCT116 colorectal cancer cells containing extra copies of chromosome 5 have a reduced proliferation rate but enhanced metastatic abilities (Stingele et al. 2012; Vasudevan et al. 2020). The observation that metastases display different CSA patterns than their primary tumors further supports that CSAs can function to drive migration or colonization (Nguyen et al. 2022). Similarly, some CSAs are known to arise at specific points in tumor evolution, such as at the transition from adenoma to carcinoma or between different disease stages (Meijer et al. 1998; Ried et al. 1999; Wistuba et al. 1999; Idbaih et al. 2008; Orsetti et al. 2014). Thus, using these correlations could give insight into which phenotype a CSA contributes to. Although aneuploidy often appears to negatively affect cell proliferation and fitness, CSAs have been implicated in angiogenesis (Brouwer et al. 2019), immune evasion (Han et al. 2021), and many other hallmarks of cancer (Zhakula-Kostadinova and Taylor 2024). Further complicating things, even the model used to study CSA can alter the outcome, with patient-derived xenografts selecting for different aneuploidies than the primary human tumor (Ben-David et al. 2017). Thus, multiple models and assays should be used to confidently tie a CSA to a phenotype.

Investigation into the role of aneuploidy in cancer thus far has proven that it is a highly complex subject dependent on several contexts. Careful selection of the modeling approach using available data sets such as The Cancer Genome Atlas (TCGA), the International Cancer Genome Consortium (ICGC), the Genomic Data Commons (GDC), and the Mitelman Database should allow for more successful studies of why aneuploidy is so ubiquitous in cancer cells (Fig. 2A).

Figure 2.

(A) The role of aneuploidy in cancer is a context-dependent subject. Tumor data sets can be used to identify the proper cellular, genetic, and phenotypic contexts in which to study aneuploidy. (B) Genetic or chemical interference with chromosome segregation can induce general aneuploidy in vitro and in vivo, whereas multiple technologies can be used to engineer specific gains or losses. Detection of general aneuploidy can be measured by quantifying DNA per cell, whereas cytogenetic and sequencing approaches allow for the determination of copy number of specific chromosomes. (C) Once representative models have been acquired, the phenotypic effects of aneuploidy can be investigated through direct assays, pathway analysis of transcriptomic data, and screening for genetic or chemical vulnerabilities. The figure was created with BioRender.com.

Engineering and detecting aneuploidy

Once a specific question has been identified and a model representative of the proper context has been selected, the aneuploidy to be studied must be engineered. As strategies for inducing aneuploidy have recently been reviewed elsewhere (Lakhani et al. 2023; Truong et al. 2023b), the most common and relevant approaches are briefly discussed here. We then summarize aneuploidy detection methods used to determine the success of aneuploidy engineering and techniques used on clinical samples (Fig. 2B).

Methods to induce general aneuploidy

As mentioned previously, aneuploidy results from CIN. Thus, a population of cells containing random aneuploidies can be created by temporary treatment with drugs that induce chromosome missegregation. Disruption of spindle geometry by brief nocodazole or monastrol treatment increases the frequency of merotelic kinetochore–microtubule (kMT) attachments and can result in lagging chromosomes (Cimini et al. 2001; Thompson and Compton 2008). However, due to SAC activation, these drugs often cause prolonged mitotic arrest that culminates in mitotic slippage or apoptosis, limiting the survival of aneuploid progeny. To bypass this arrest, the SAC can be disrupted directly by treatment with MPS1 inhibitors. Such inhibition allows for the entry into anaphase regardless of the presence of misaligned chromosomes and improper kMT attachments, leading to missegregation and high levels of aneuploidy (Santaguida et al. 2010). It should be noted that these methods often lead to increased DNA damage due to unresolved chromatin bridges or chromothripsis in damaged micronuclei after chromosome missegregation (Dalton et al. 2007; Quignon et al. 2007; Crasta et al. 2012; Zhang et al. 2015). To induce WGD, cytokinesis can be blocked through treatment with dihydrocytochalasin B or D. Due to the toxicity of aneuploidy-inducing drugs, genetically engineered mouse models (GEMMS) are more commonly used for in vivo studies. Models of CIN-induced tumorigenesis often include altering SAC function by modifying the expression of proteins like MAD2 or BUB1 or generating supernumerary centrosomes via PLK4 overexpression (Holland and Cleveland 2009; Schvartzman et al. 2010). To prevent lethality and allow for tissue-specific studies, inducible and conditional models have been created (Foijer et al. 2017; Hoevenaar et al. 2020; Trakala et al. 2021).

Methods to generate chromosome-specific aneuploidy

Methods to engineer CSA can be divided into those that induce gains, losses, or both. Traditionally, microcell-mediated chromosome transfer (MMCT) has been used to introduce and interrogate chromosome-specific gains in vitro. Injection of mouse cells after MMCT of human chromosomes has been used to identify tumor and metastasis suppressors, but applications with living organisms are limited, as the MMCT protocol must be performed in vitro (Doherty and Fisher 2003). Additional limitations of this method are that chromosomes trapped in micronuclei are prone to chromothripsis, and micronuclei themselves have been found to activate metastasis-inducing cGAS/STING signaling (Bakhoum et al. 2018; Kneissig et al. 2019). Thus, MMCT models of aneuploidy may display phenotypes caused by the method itself rather than the gain of the chromosome being studied.

There are two main approaches for generating cells with specific whole-chromosome losses. One is cre/loxP-mediated chromosomal elimination, which generates acentric and dicentric chromatids that undergo missegregation (Lewandoski and Martin 1997). The downside to this approach is that it requires prior genetic engineering to insert the loxP sites in order to induce chromosome loss. The second method for chromosome deletion is nuclease (often Cas9)-mediated chromosome fragmentation. This is accomplished by inducing double-strand breaks (DSBs) along the length of the chromosome or near the centromere (Adikusuma et al. 2017; Zuo et al. 2017). This method can also be used to engineer stable arm-level chromosomal losses when pericentromeric Cas9-targeted cells are supplied with a donor sequence containing homology with the targeted region followed by an artificial telomere sequence (Uno et al. 2017; Taylor et al. 2018). Furthermore, enrichment for cells lacking the targeted chromosome can be achieved using a negative selection cassette inserted at the region of interest, as demonstrated in the ReDACT approach (Girish et al. 2023). Caveats of Cas9-mediated chromosomal deletion include increased DNA damage due to off-target effects and potential recombination of the generated chromosomal fragments. Finally, multiple nuclease-dead Cas9 (dCas9)-mediated technologies that cause targeted chromosome missegregation have been developed in the past few years, allowing for the generation of both segmental and whole gains and losses of specific chromosomes. These approaches involve targeting a mutant kinetochore protein to the centromere, such as in the KaryoCreate approach (Bosco et al. 2023), or targeting a minus end-directed kinesin (Truong et al. 2023a) or ectopic kinetochore formation (Tovini et al. 2023) to other chromosome-specific repetitive sequences.

All of the approaches discussed above are used to induce CSA in a population of cells from which single-cell-derived clones are then picked and analyzed for the effects of the CSA. This approach comes with the caveats that it is labor-intensive, limited in genetic backgrounds, and subject to adaptation. Because the vast majority of cancer cell lines display genetic instability, genetic heterogeneity within the population is present and constantly being generated. By choosing a handful of clones to analyze, the genetic context of the CSA is extremely limited and may lead to unrepresentative associations between a CSA and a certain phenotype. Additionally, expanding clones from single cells can take months, during which genomically unstable cancer cell lines could mutate to adapt to or reverse the engineered aneuploidy (Girish et al. 2023; Bökenkamp et al. 2025). A less targeted yet unbiased approach to study CSA would be to induce random aneuploidy in a population and observe which CSAs are selected for under different conditions. This forward genetics approach has already been used to uncover aneuploidies that confer resistance to certain drugs (Ippolito et al. 2021; Lukow et al. 2021). Combination of these aneuploidy selection screens with analysis of available cancer data sets could aid in identifying and distinguishing driver gains and losses from passenger CSAs.

It is important to note that engineering CSA in mouse models for cancer studies is not a common practice due to several technical and biological limitations. First, the use of MMCT to introduce whole-chromosome gains can only be performed in vitro. Thus, cells containing chromosome gains either are injected into blastocysts to generate GEMMs or are used in allograft experiments to assay the effect of a CSA on tumor growth (Zhang et al. 2016; Kazuki et al. 2020). Neither of these is an ideal model for studying how CSA contributes to tumorigenesis, as germline gains are often embryonic-lethal (Epstein 1985) and allograft experiments are not representative of the evolution and microenvironment of spontaneous tumorigenesis. Although methods to generate conditional, tissue-specific chromosome loss have been developed, their application to model tumorigenesis has been limited (Grégoire and Kmita 2008; Tada et al. 2009). One approach that has been used to study the tissue-specific effects of aneuploidy in vivo involves the ablation and subsequent reconstitution of bone marrow with aneuploid hematopoietic stem cells (Pfau and Amon 2015; Sano et al. 2022). Although technically challenging, this approach could be expanded on to generate mice with aneuploidy only in a certain organ via blastocyst complementation or other transplantation models (Heyer et al. 2010; Founta and Papanayotou 2022). Even if CSA could be easily engineered in vivo, the ability to directly model the effects of CSAs observed in human tumors would be limited due to the incomplete synteny between the human and mouse genomes. Because each human chromosome maps to segments on multiple mouse chromosomes, recapitulating the gain or loss of an entire human chromosome would require coordinated alterations across several mouse chromosomes, necessitating complex and extensive genetic engineering. An alternative approach could be the construction of a mouse artificial chromosome carrying the mouse homologs of genes found on a single human chromosome. This strategy could be used to investigate how the collective gene content of a specific human chromosome contributes to tumorigenesis in vivo (Oshimura et al. 2015; Uno et al. 2022).

Methods to detect general aneuploidy

Various approaches can be used to detect and determine the overall aneuploidy level in cells. The oldest approach is to perform metaphase spreads and subsequent chromosome counting with or without the help of techniques such as multicolor fluorescent in situ hybridization (mFISH) or spectral karyotyping (SKY) (Bakker et al. 2015). In addition to being labor-intensive and limited to dividing cells, this approach can be error-prone, as metaphases from multiple cells can overlap, making it difficult to assign chromosomes to individual cells and therefore accurately quantify chromosome number per cell. A simpler and high-throughput approach to characterize the aneuploidy of a population is to quantify the amount of DNA per cell using flow cytometry of cells stained with DNA markers such as propidium iodide or DAPI (Danielsen et al. 2016). Although lower resolution than karyotyping methods, cytometric methods do not require dividing cells and thus can be used with patient samples. The methods described below to detect CSA can also be used to quantify overall aneuploidy by summing the number of arm-level or whole-chromosome-level gains and losses. As CIN has been a significant subject of this review, it is relevant to mention that CIN quantification is commonly accomplished through immunofluorescence and live-cell imaging of lagging chromosomes, anaphase bridges, and micronuclei or inferred through heterogeneity analysis from single-cell sequencing (Lynch et al. 2024).

Methods to detect chromosome-specific aneuploidy

The methods available to detect CSA have expanded and become higher resolution over time. FISH on metaphase spreads from dividing cells or on interphase nuclei from nondividing cells can allow for the visualization of specific loci or centromeres if using human samples. Quantification of FISH foci or labeled chromosomes from mFISH or SKY per nucleus can be used to determine the frequency of a CSA in a population of cells. Furthermore, a novel variation of FISH in suspension and subsequent imaging with cytometers capable of sensitive imaging allow for the combination of the specificity of FISH with the high throughput of flow cytometry (Minderman et al. 2012; Worrall et al. 2018). Other than direct imaging of chromosomal loci, several approaches to detect CSA involve determining the relative abundance of a region of interest compared with that of a control region. This strategy is more ideal for the characterization of clones than detection of aneuploidy within a population due to the heterogeneity of populations affecting signal strengths. The lowest-resolution version of this approach is to perform qPCR for multiple loci across the arm or whole chromosome of interest (Mayo et al. 2010). Array-based comparative genome hybridization (Shinawi and Cheung 2008) and low-pass whole-genome sequencing (Shen and Seshan 2016; Wang et al. 2016) can also be used to directly measure chromosomal imbalances, whereas RNA sequencing (Bařinka et al. 2022) data can be used to infer copy number alterations. Finally, copy number calling from single-cell data can be used for high-throughput detection of subclonal CSAs within a sample. Due to the need to amplify the entire genome, single-cell DNA sequencing technology is relatively low-throughput, high-cost, and therefore currently less developed and applied for aneuploidy detection than single-cell RNA sequencing technology (Mallory et al. 2020). On the other hand, single-cell RNA sequencing is limited in that it relies on inference from RNA levels, with a low proportion of expressed genes in the genome making only low-resolution copy number calls trustworthy (Chen et al. 2025). Therefore, DNA sequencing is more reliable for the detection of small copy number variations. However, single-cell RNA sequencing presents the unique opportunity to detect ploidy levels and transcriptome profiles at the same time, allowing for the immediate linking of aneuploidy to its effects on gene expression (Mahdipour-Shirayeh et al. 2022). The continued development of these single-cell technologies will allow for more comprehensive and precise characterization of tumor heterogeneity, enabling improved understanding of how chromosomal alterations drive cancer evolution.

Recent findings on the contribution of aneuploidy to cancer

Thus far, we have established that the role of aneuploidy in cancer is a context-dependent subject that requires the refinement of variables for clearer elucidation. Because studies investigating this relationship have been recently summarized elsewhere (Hosea et al. 2024; Klockner and Campbell 2024; Zhakula-Kostadinova and Taylor 2024), below we demonstrate how the computational analyses and aneuploidy engineering and detection technologies discussed thus far have recently been applied to successfully tie CSA to cancer phenotypes (see “Aneuploidy and Cell-Autonomous Cancer Phenotypes: 4p Loss and 1q Gain”). We then highlight how both general and specific forms of aneuploidy contribute to two processes highly relevant to patient survival: immune evasion and metastasis (see “Aneuploidy and Non-Cell-Autonomous Cancer Phenotypes”).

Aneuploidy and cell-autonomous cancer phenotypes: 4p loss and 1q gain

The goal of the first section of this review was to argue that cell, genetic, and phenotypic contexts can influence the apparent relationship between aneuploidy and cancer. The ultimate purpose of this discussion was to emphasize the importance of diligently selecting a particular aneuploidy–cancer relationship and relevant models for studying it. We then presented aneuploidy engineering and detection methods that could be used to generate said ideal models. Here, we present two recent real-world examples of how computational analyses and chromosome-engineering strategies can be used to identify and demonstrate causal relationships between aneuploidy and cancer-relevant phenotypes, focusing here on cell-autonomous cancer phenotypes.

Although general aneuploidy is normally considered detrimental for cellular proliferation and cellular fitness, CSA can promote proliferation and tumorigenesis. The current working hypothesis of how CSA serves as a cancer driver is by altering the dosage of tumor suppressors and oncogenes. The study by Kuzmin et al. (2024) uses many of the strategies discussed in this review to support this hypothesis, demonstrating that 4p loss enhances the proliferation of breast cancer cells by lowering the dosage of a tumor suppressor, C4of19. First, the investigators established the relevance of investigating 4p loss in TNBC by narrowing the genetic context to basal breast cancers, which typically exhibit deletions (Curtis et al. 2012). They established that 4p loss is common (∼65%) in this context and that it predicts poor prognosis independent of p53 status, eliminating the confounding variable of CIN. Gene set enrichment analysis on TCGA transcriptomic data between samples with and without the chromosome 4p loss suggested that samples with 4p loss may hold a proliferative advantage. Additionally, single-cell RNA-seq was performed to simultaneously infer copy number and evaluate gene expression patterns, which revealed that cells lacking 4p loss were depleted in proliferation markers. Phylogenetic reconstruction of patient-derived samples indicated that 4p loss is a preferential early event. Finally, an overexpression screen was performed in multiple patient-derived cell lines to identify tumor suppressors on chromosome 4p. Interestingly, only some cell lines showed slowed growth upon overexpression of 4p genes. The investigators hypothesized that this could be because 4p loss interacts with additional genetic events only present in some of the cell lines studied to confer a proliferative advantage. Importantly, this illustrates the drawback of using different cell lines to study the effect of aneuploidy rather than engineered isogenic models.

The second example study by Girish et al. (2023) avoided the confounding variable of differing genetic backgrounds by engineering isogenic clones with or without multiple CSAs, focusing mainly on the gain of chromosome arm 1q. First, the investigators used timing analysis on patient whole-genome sequencing data to identify chromosome 1q gain as one of the first copy number alterations to occur in melanoma and breast cancer. Additionally, they performed survival analysis showing 1q gain to be a poor prognostic factor and mutual exclusivity analysis between arm gains and mutations that found chromosome 1q gain to be mutually exclusive with mutations in TP53. In order to experimentally study the effects of 1q gain, the investigators used several CRISPR-based strategies to reverse chromosome 1q trisomy in melanoma, gastric cancer, and ovarian cancer cell lines. Their strategy, ReDACT, consisted of cutting proximal to the centromere in order to induce 1q loss. The resulting isogenic disomic and trisomic clones were assayed for doubling time, anchorage-independent growth, and tumor formation in mice, with the results leading to the conclusion that cells with chromosome 1q gain are dependent on this CSA to maintain malignant growth. Subsequent genetic screening experiments led to the mechanistic explanation that the gain of chromosome 1q promotes survival by increasing the expression of the p53 suppressor MDM4, which is encoded on 1q. Finally, the investigators used their isogenic clones to identify a therapeutic vulnerability of 1q gain, highlighting aneuploidy dependency as a potential therapeutic target in specific cancer types.

Together, the summaries above demonstrate how researchers in the aneuploidy field are beginning to take advantage of abundant tumor data and genome-engineering technologies to successfully identify the mechanisms behind why CSAs occur in human tumors.

Aneuploidy and non-cell-autonomous cancer phenotypes

In addition to promoting cell-intrinsic cancer phenotypes, aneuploidy can profoundly influence non-cell-autonomous tumor phenotypes. Here, we discuss recent findings showing how both general and chromosome-specific aneuploidy shape the tumor microenvironment, promoting immune evasion, resistance to immunotherapy, and metastasis.

General aneuploidy and the tumor microenvironment

As discussed previously, aneuploidy levels can predict overall survival. Additionally, copy number alteration (CNA) burden is correlated with recurrence, time to treatment failure, and other clinical outcomes (Hieronymus et al. 2018; Van Den Bosch et al. 2022). Recent studies show that the interaction between aneuploidy and the microenvironment can influence tumor metastasis, immune evasion, and response to immunotherapies.

General aneuploidy has been widely implicated in immune evasion and resistance to immune checkpoint inhibitors (ICIs) in human cancers. Tumors of multiple types with high levels of aneuploidy exhibit reduced cytotoxic immune infiltration, impaired antigen presentation, and immunosuppressive tumor microenvironments (TMEs), leading to diminished responses to ICI therapy. In melanoma and non-small cell lung cancer (NSCLC) patient cohorts treated with anti-CTLA-4 or anti-PD-1/PD-L1 inhibitors, high tumor aneuploidy was significantly associated with lower overall survival and reduced therapeutic efficacy (Davoli et al. 2017; Tamborero et al. 2018). Similarly, retrospective analyses of NSCLC cohorts showed that tumors with high aneuploidy exhibited lower CD8⁺ T-cell infiltration and reduced expression of immune activation markers, which correlated with worse clinical outcomes following ICI treatment (Frigola et al. 2022; Alessi et al. 2023). Similar trends have been observed in gastrointestinal cancers, where low CNA burden correlated with better responses to ICIs, whereas highly aneuploid tumors showed resistance to checkpoint blockade therapy (Lu et al. 2020). These findings suggest that assessing tumor aneuploidy levels could help stratify patients for ICI treatment, improving precision immunotherapy approaches.

The reduced immune infiltration and lower responsiveness of aneuploid tumors to ICI therapy indicate that chromosomal alterations may actively suppress antitumor immunity. In one study using in vitro model systems of aneuploidy and coculture with immune cells, aneuploidy was shown to induce proteotoxic stress, leading to activation of the unfolded protein response (UPR) as a compensatory mechanism to manage misfolded proteins. In turn, UPR activation alters cytokine secretion, leading to the dysregulation of macrophages and T cells. This shift in the TME to limited CD8⁺ T-cell infiltration creates a protumorigenic and immunosuppressive niche that contributes to immune evasion and resistance to ICIs (Xian et al. 2021). CIN has also been found to influence the TME through cytosolic DNA inducing chronic activation of the cGAS–STING pathway. Although STING signaling is typically associated with antitumor immunity, persistent CIN-induced activation leads to NFkB signaling and resultant immune suppression, angiogenesis, and metastasis (Bakhoum and Cantley 2018; Bakhoum et al. 2018; Kuang and Li 2022; Li et al. 2023). Although the mechanism is yet to be elucidated, STING activation in CIN tumor cells has been linked to M2 macrophage polarization and the recruitment of myeloid-derived suppressor cells (Liang et al. 2017; Li et al. 2023). Additionally, chronic inflammatory signaling results in the release of IL-6, which promotes tumor cell survival and PD-L1 expression to further contribute to immune evasion (Hong et al. 2019; Vasiyani et al. 2022; Kandala et al. 2023).These findings suggest that targeting the UPR and STING pathways may reduce immune-evasive TME editing and enhance the success of ICIs in aneuploid tumors. Additionally, they present a mechanism of CIN driving malignancy beyond the introduction of karyotypic heterogeneity.

It should be noted that the interaction between chromosomal aneuploidy and CIN with the tumor immune microenvironment remains an area of active investigation. For instance, a recent study demonstrated that acute induction of chromosomal instability through MPS1 inhibition can promote macrophage polarization toward antitumor phenotypes (Hayes et al. 2024)—findings that are in contrast to those reported by Xian et al. (2021). These divergent conclusions are likely attributable to differences in model systems and MPS1 inhibitor concentrations. Nevertheless, the apparent contradiction underscores the need for more comprehensive and systematic studies to achieve a deeper and more unified understanding of this complex process.

Chromosome-specific aneuploidy and immune evasion: a prominent role for 9p loss

Chromosomal deletions play a critical role in shaping the tumor microenvironment and determining immune evasion mechanisms. Among these, chromosome 9p loss has emerged as a key driver of resistance to immune checkpoint therapy (ICT), particularly in NSCLC, head and neck squamous cell carcinoma, and pancreatic cancer. Recent studies have identified that loss of 9p, or segments thereof, disrupts essential immune regulatory pathways, leading to an immune cold tumor phenotype and impaired response to PD-L1 blockade (Alessi et al. 2023). This deletion is strongly associated with low levels of key immune-related genes such as the type I IFN gene cluster, JAK2, and CDKN2A, which are encoded on 9p and are crucial for recruiting CD8⁺ T cells and sustaining an interferon-driven antitumor immune response (Fig. 3A; William et al. 2021; Barriga 2023).

Figure 3.

(A, left panel) Schematic representation of chromosome 9p showing known key genes with immune functions playing critical roles in the tumor microenvironment. (Right panel) Schematic representation of the decrease in infiltrating CD8⁺ cells that is found in tumors with 9p deletion as compared with tumors without it. (B) Schematic representation of how 9p loss may contribute to cancer immune evasion. Decreased gene dosage of key genes such as JAK2 and PD-L1 and the type I IFN gene cluster (including IFNɛ), along with CDKN2A, in tumor cells with 9p loss results in reduced levels of proinflammatory cytokines such as IFNɛ and CXCL9/10. This reduction leads to decreased activation of dendritic cells and macrophages, ultimately resulting in less recruitment of CD8⁺ T cells to the tumor microenvironment. The figure was created with BioRender.com.

A major consequence of 9p loss is the depletion of CXCL9 and CXCL10 in the tumor microenvironment (normally produced by dendritic cells and macrophages), resulting in diminished CD8⁺ T-cell infiltration and a suppressed immune landscape. This effect is mediated through loss of type I interferons, particularly IFNɛ, which is predominantly expressed in tumor cells and contributes to innate immune activation. In murine models with syntenic 9p deletions, tumors exhibited a marked reduction in CXCL9/10-expressing macrophages and dendritic cells, confirming the causal link between 9p loss and immune suppression (Barriga 2023). Notably, clinical studies have shown that 9p loss correlates with resistance to ICT independent of tumor mutation burden or PD-L1 expression. Large-scale analyses from TCGA demonstrate a strong negative correlation between 9p deletion and CD8⁺ T-cell infiltration across multiple tumor types (Han et al. 2021).

The immunosuppressive effects of 9p loss extend beyond interferon depletion. Deletion of JAK2 at 9p24.1 disrupts interferon γ signaling, further reducing CXCL9/10 expression and reinforcing an immune-excluded tumor state. Additionally, PD-L1 loss due to 9p deletion suggests that these tumors are inherently resistant to ICT, as the primary target of PD-1 blockade is absent (Fig. 3B; Li et al. 2023). These findings highlight 9p loss as a potential predictive biomarker for ICT resistance and suggest that therapies designed to restore CXCL9/10 expression or enhance interferon signaling could improve responses in these patients.

Several therapeutic strategies have been proposed to counteract the effects of 9p loss. One promising approach involves in situ vaccination using CXCL9/10 engineered dendritic cells, which has shown efficacy in murine models by restoring CD8⁺ T-cell recruitment and improving ICT response (Lim et al. 2024). Alternative strategies include STING agonists to activate innate immune-sensing pathways and JAK2 modulation to restore interferon responsiveness. As chromosomal deletions become increasingly recognized as determinants of immune evasion, further studies on 9p loss-directed therapies could lead to new precision immunotherapy approaches for patients with ICT-resistant tumors. Notably, despite 9p loss being associated with immune cold phenotypes across several tumor types, this is not universal. For example, in renal cancer, 9p loss is associated with poorer survival, but this does not seem to be due to a decrease in the immune infiltrate, as it actually appears to be higher in tumors with 9p loss compared with others (Braun et al. 2020) This highlights the tissue specificity of how CSAs shape the tumor microenvironment and antitumor immune responses and the need to consider these contextual specificities in designing tumor type-specific immunotherapy strategies.

Chromosome-specific aneuploidy and metastasis

Metastatic tumors frequently exhibit selective chromosomal gains and losses that provide advantages for survival, invasion, and adaptation to new microenvironments. Among the most recurrent alterations, chromosome 8q gain (8q24, including MYC) is commonly observed in colorectal cancer (CRC) and breast cancer, where it enhances proliferation, survival, and metabolic reprogramming. This alteration is particularly enriched in liver metastases of CRC and has been associated with increased metastatic aggressiveness (Joosse et al. 2018; Winnard et al. 2020; Negoto et al. 2022; Nguyen et al. 2022). Similarly, amplification of chromosome 20q is a hallmark of metastatic CRC and gastric cancer. This region contains AURKA and TPX2, which promote tumor progression and chromosomal instability, making 20q gain one of the most frequent aberrations detected in CRC liver metastases (Joosse et al. 2018; Winnard et al. 2020; Negoto et al. 2022).

Other notable changes include chromosome 6p gain, identified in CRC and melanoma metastases, which enhances cell invasion and differentiation via TFAP2C regulation (Winnard et al. 2020). In CRC and breast cancers, chromosome 1q gain has been linked to therapy resistance and survival advantages through amplification of MCL1, a gene that inhibits apoptosis and facilitates metastatic persistence (Winnard et al. 2020; Nguyen et al. 2022). Losses of tumor suppressor loci also play a critical role. Chromosome 16p loss, particularly in CRC, has been associated with increased tumor invasiveness and the ability of cancer cells to intravasate into circulation, favoring metastatic dissemination (Joosse et al. 2018). Similarly, deletion of chromosome 10q, which includes PTEN, disrupts cell cycle control and the PI3K–AKT pathway, supporting metastatic progression in melanoma and CRC (Winnard et al. 2020).

Overall, these alterations reflect selective pressures acting on metastatic clones, stabilizing specific chromosomal changes that enhance adaptability. Although the degree of CIN varies across tumor types and metastatic sites, these recurrent aneuploidies provide functional insights into metastatic evolution and potential therapeutic vulnerabilities.

Therapeutic strategies for targeting aneuploidy in cancer

General aneuploidy-induced vulnerabilities

General aneuploidy creates cellular stresses that cancer cells must counteract to survive, as described above in “Contextual Considerations for the Study of Aneuploidy in Cancer.” These stresses, including mitotic instability, proteotoxic burden, and metabolic imbalances, form the basis for synthetic-lethal interactions that allow for the selective targeting of aneuploid cells (Fig. 4). Here we discuss three examples of vulnerabilities associated with the aneuploid state. An additional type of aneuploidy-associated vulnerability (not discussed here in detail) exploited by some types of conventional anticancer chemotherapies (e.g., microtubule poisons) is CIN itself. Because cancer cells tolerate only limited aneuploidy, driving excessive chromosome errors with chemotherapy preferentially kills malignant cells (Bhatia et al. 2024).

Figure 4.

(A) The increased amount of genetic material and genetic imbalances induced by general aneuploidy confer on cells vulnerabilities that can be specifically targeted. Pushing the level of CIN in already unstable cells by interfering with SAC signaling or spindle dynamics is one proposed approach for targeting CIN cancer cells. Additionally, the stoichiometric imbalances of RNA and proteins put stress on degradation machinery, making aneuploid cells more sensitive to inhibition of these processes. Finally, aneuploidy and CIN are associated with higher response to radiation therapy but lower response to ICI, and new potential therapeutic strategies are represented here. (B) The gain or loss of specific genes due to CSA generates unique weaknesses in the aneuploid cells. Deletions increase dependency on paralogous genes and can allow for targeting of haploinsufficiencies, whereas gains can increase the metabolism of prodrugs into their toxic forms.

Sensitivity of aneuploid cells to SAC and KIF18A inhibition

The SAC is a crucial mitotic surveillance mechanism that ensures accurate chromosome segregation by delaying anaphase onset until all chromosomes achieve proper attachment to the mitotic spindle. SAC activation prevents premature mitotic exit by inhibiting the anaphase-promoting complex/cyclosome (APC/C), thus safeguarding genomic stability. Defects in SAC function can lead to chromosome missegregation, aneuploidy, and tumorigenesis. As discussed above in “Contextual Considerations for the Study of Aneuploidy in Cancer,” high levels of CIN are toxic to cells. Stemming from this is the idea that increasing CIN in already aneuploid and unstable tumors could be a successful strategy to selectively eliminate cancer cells. Recent studies indicate that aneuploid cells exhibit increased sensitivity to inhibition of the SAC, the mitotic kinesin KIF18A, and mitotic kinases, making these attractive therapeutic targets in cancer (Fig. 4A).

In a study that analyzed large-scale genetic and chemical perturbation screens, Cohen-Sharir et al. (2021) demonstrated that aneuploid cancer cells exhibit increased dependency on SAC core components, such as BUB1B, MAD2, and TTK. Their study found that although aneuploid cells can initially bypass SAC inhibition more readily than diploid cells, upon prolonged treatment, they ultimately suffer from severe viability and proliferation defects due to increased CIN. Upon further investigation, they found that mRNA levels for the microtubule-based motor protein KIF18A were reduced, and sensitivity to KIF18A knockdown was greater in their aneuploid cells than in diploid cells. This sensitivity correlated with higher markers of CIN, such as multipolar spindles, mitotic delay, and micronuclei in KIF18A knockdown aneuploid cells but not in diploid cells (Cohen-Sharir et al. 2021; Marquis et al. 2021). One proposed mechanism for the synthetic lethality of KIF18A inhibition and CIN is that CIN cells have increased microtubule dynamics and therefore require more KIF18A activity to stabilize kMT attachments (Cohen-Sharir et al. 2021; Marquis et al. 2021). The inability to satisfy the SAC induced by KIF18A inhibition causes extended SAC activation that ultimately leads to death in mitosis or lethal segregation errors (Gliech et al. 2024). Interestingly, WGD⁺ cells are particularly dependent on KIF18A for proper mitotic function (Quinton et al. 2021). This could be because WGD⁺ cells already take longer to satisfy the SAC due to the increased number of proper kMT attachments to be made (Yang et al. 2008). In conclusion, the ability of KIF18A inhibition to selectively increase SAC activity and push CIN cells out of a tolerable range of instability presents as a promising therapy for the targeted killing of cancer cells.

In addition to SAC proteins and KIF18A, previous studies have suggested that aneuploid cells are more sensitive to Aurora A and Aurora B inhibition due to their increased reliance on these kinases for mitotic progression in the presence of extra chromosomes and centrosomes. Aurora A plays a critical role in centrosome maturation and spindle assembly, whereas Aurora B ensures accurate chromosome segregation and cytokinesis. Although Aurora kinase inhibitors have yet to achieve significant clinical success, efforts to develop and test more effective inhibitors are ongoing (Du et al. 2021).

Current evidence indicates that small molecule strategies targeting aneuploidy—particularly Aurora kinase inhibitors and, more recently, inhibitors of the SAC and KIF18A—demonstrate strong mechanistic selectivity for aneuploid cells in preclinical models of both solid and hematologic cancers. However, clinical translation has been hindered so far by several factors. For example, Aurora kinase inhibitors have generally shown only modest single-agent efficacy and have been limited by dose-limiting toxicities, including hematologic and gastrointestinal adverse events, in early phase trials (Falchook et al. 2015; Goldenson and Crispino 2015; Niu et al. 2015). Clinical adoption has been further hampered by the lack of predictive biomarkers (e.g., aneuploidy load), the absence of patient stratification, and the development of drug resistance through mechanisms such as kinase domain mutations and compensatory signaling pathway upregulation (Goldenson and Crispino 2015). SAC and KIF18A inhibitors demonstrate synthetic-lethal effects in high-aneuploidy cancer cell lines and show strong preclinical tumor selectivity, but clinical data are largely lacking, with questions remaining about long-term efficacy, safety, and optimal patient selection strategies (Cohen-Sharir and Ben-David 2021; Cohen-Sharir et al. 2021). Recent efforts have increasingly focused on genomic stratification and combination regimens to overcome resistance, but durable and safe clinical translation of these approaches has yet to be achieved (Falchook et al. 2015; Goldenson and Crispino 2015; Niu et al. 2015; Cohen-Sharir and Ben-David 2021; Cohen-Sharir et al. 2021).

Relationship between general aneuploidy and radiation or immune checkpoint inhibitor therapy

Patients with high aneuploidy or CIN and low tumor mutation burden may benefit from combination therapies, such as radiotherapy plus ICI. Moreover, general aneuploidy plays a role in radiotherapy sensitivity, where high-aneuploidy tumors demonstrate enhanced responses to concurrent radiotherapy and ICI, supporting the notion of treatment intensification. In the neoadjuvant and adjuvant settings, aneuploidy score can be used to guide treatment intensification in locally advanced cancers, including lung, breast, and gastrointestinal cancers (Cosper et al. 2022; Spurr et al. 2022). Furthermore, CIN also provokes innate immune signaling within tumors that may be targeted. Hong et al. (2022) demonstrated that CIN activates the cGAS–STING pathway through cytosolic micronucleus formation, driving an IL-6/STAT3-dependent inflammatory loop that supports tumor cell survival. Notably, blockade of this pathway, such as with the IL-6R antibody tocilizumab, selectively impairs CIN-high tumor growth, revealing a novel vulnerability in chromosomally unstable cancers (Hong et al. 2022).

Dependence on enhanced RNA and protein degradation

Aneuploidy disrupts the stoichiometric balance of gene expression, leading to cellular stress. The transcriptional burden in aneuploid cells results from the overexpression of genes due to extra chromosomal material. To mitigate this imbalance, aneuploid cells adapt by enhancing RNA and protein degradation pathways, creating specific vulnerabilities that can be therapeutically targeted. In addition, aneuploid cells upregulate RNA degradation mechanisms, including nonsense-mediated decay (NMD) and microRNA-mediated silencing. This heightened RNA turnover helps maintain cellular homeostasis by preventing the accumulation of aberrant transcripts. However, it also renders aneuploid cells more susceptible to disruptions in RNA degradation pathways. Experimental evidence indicates that inhibiting key components of these pathways selectively impairs the viability of aneuploid cells, suggesting potential therapeutic strategies (Ippolito et al. 2024). At the protein level, the imbalance caused by aneuploidy leads to proteotoxic stress due to the disproportionate production of proteins, especially disrupting the stoichiometry between the subunits of protein complexes (Oromendia and Amon 2014). Aneuploid cells counteract this by reducing protein synthesis and increasing degradation of overabundant proteins primarily through the ubiquitin–proteasome system. This adaptation is crucial for their survival but also makes them particularly sensitive to proteasome inhibitors. Clinical observations have shown that multiple myeloma patients exhibit a favorable response to proteasome inhibition therapy, highlighting a viable treatment avenue (Ippolito et al. 2024).

Dependence on metabolic pathways

A recent study identified possible metabolic vulnerabilities associated with the aneuploid state (Magesh et al. 2025). Aneuploid cells face metabolic stress due to imbalances in nucleotide and energy production. CRISPR screens and multiomics analyses comparing isogenic diploid and aneuploid immortalized human cells identified that aneuploid cells rely heavily on pyrimidine biosynthesis and mitochondrial oxidative phosphorylation for survival. This dependency arises from insufficient nucleotide pools, resulting in disrupted DNA replication and S-phase arrest when these pathways are inhibited. Loss of pyrimidine synthesis or impaired mitochondrial metabolism leads to synthetic lethality in aneuploid cancer cells, distinguishing them from diploid cells.

Chromosome-specific aneuploidy-induced vulnerabilities

In addition to general aneuploidy vulnerabilities, specific chromosomal losses or gains create unique vulnerabilities related to the imbalanced dosage of specific genes encoded on the altered chromosomes. These imbalances have led to the identification of synthetic-lethal relationships between certain genetic features and arm-level aneuploidies (Fig. 4B; Shukla et al. 2020).

An effective approach to cancer treatment is targeting amplified oncogenes, such as ERBB2. However, these amplifications typically involve dozens to >100 copies per cell. In contrast, aneuploidy-driven copy number changes usually occur at a lower magnitude, often by onefold or twofold, and have not been widely successful as therapeutic targets. A more promising approach to targeting aneuploidy is known as collateral vulnerability. This concept refers to the exploitation of passenger gene deletions to develop cancer-selective therapies by leveraging cancer cells’ dependency on homologous genes, exacerbating haploinsufficiency, or targeting alternative compensatory pathways (Muller et al. 2015). Below, we describe three key examples of collateral vulnerabilities associated with cancer aneuploidy. Although collateral vulnerabilities most commonly arise from deletions, we also present an example in which a chromosome gain increases susceptibility to a prodrug by enhancing its conversion into its active form.

Lenalidomide: an aneuploidy-targeting drug used in the clinic

Lenalidomide is an approved targeted therapy for the treatment of multiple myeloma and myelodysplastic syndrome (MDS) with deletion of chromosome 5q. Lenalidomide exerts its function through modulation of the cereblon (CRBN) E3 ubiquitin ligase complex, altering its substrate specificity to promote the ubiquitination and proteasomal degradation of different proteins including CK1α (Krönke et al. 2015). CK1α is a negative regulator of p53 and Wnt signaling encoded by CSNK1A1 on 5q. Although CK1α haploinsufficiency increases Β-catenin-induced expansion of hematopoietic stem cells, total ablation of CK1α activity leads to p53-mediated apoptosis (Järås et al. 2014; Schneider et al. 2014). Thus, lenalidomide allows for the targeted killing of hyperproliferative cells in MDS patients by exacerbating CK1α haploinsufficiency.

Dependency on paralog genes due to gene deletion

Passenger deletions in cancer can create therapeutic vulnerabilities by eliminating functionally redundant genes essential for tumor survival. Paralog genes are genes that share a high degree of similarity and perform often redundant functions. Deletion in an essential gene, especially when homozygous, would lead to a dependence of the cell on the other paralog genes for the essential function. A notable example is the deletion of ENO1 in glioblastoma, which removes the key glycolytic enzyme Enolase 1. Tumor cells exhibiting this loss rely on the homologous enzyme Enolase 2 for glycolysis. Targeting Enolase 2 selectively impairs the viability of ENO1-deleted glioblastoma cells while sparing normal cells (Muller et al. 2012)

Although homozygous deletions provide the strongest rationale for this therapeutic approach by preventing resistance through gene restoration, heterozygous deletions can also induce similar paralog dependencies. A key example is 8p loss, one of the most common aneuploidies in cancer. In one study, the investigators experimentally modeled a specific ∼30 Mb deletion on chromosome 8p in hepatocellular carcinoma cell lines, generating isogenic cell lines with and without the deletion. The study demonstrated that 8p deletions lead to the upregulation of genes that promote metastasis. Importantly, the loss of specific genes on 8p creates unique vulnerabilities in cancer cells, which can be exploited therapeutically. One key vulnerability identified is the dependency on Nudix hydrolase 17 (NUDT17), an enzyme involved in nucleotide metabolism. The study demonstrated that the loss of NUDT18, a paralog located on 8p, renders cancer cells highly dependent on NUDT17 for maintaining nucleotide homeostasis. Inhibiting NUDT17 selectively impairs the survival of 8p-deleted cancer cells, highlighting a potential synthetic-lethal target for tumors harboring 8p loss (Huth et al. 2023). By identifying synthetic-lethal interactions specific to 8p-deleted tumors, the study suggests potential personalized treatment strategies for patients with this genetic alteration.

In another study, the investigators performed computational analysis using the DepMap data set to identify vulnerabilities associated with 8p loss (Krieg et al. 2024). Through an integrative analysis of large-scale cancer genomics and functional data sets, they identified the mitochondrial iron transporter MFRN2 as a specific vulnerability in tumors harboring 8p deletions. The study revealed that loss of MFRN2 in cancer cells deficient in its paralog, SLC25A37 (encoding MFRN1), which resides on chromosome 8p, leads to impaired mitochondrial respiration, depletion of iron–sulfur cluster proteins, DNA damage, and cell death. In vivo experiments demonstrated that targeting MFRN2 in MFN1-deficient tumors significantly hindered tumor growth and, in some cases, resulted in complete tumor eradication. These findings suggest that MFRN2 is a promising therapeutic target for cancers with chromosome 8p deletions, with MFRN1 expression serving as a potential biomarker for patient stratification.

CYCLOPS genes: heterozygous deletions exacerbating haploinsufficiency

CYCLOPS (copy number alterations yielding cancer liabilities owing to partial loss) genes are single-copy essential genes found on partially lost chromosomes (Nijhawan et al. 2012). Cancer cells are highly vulnerable to perturbations of these genes, as they lack a redundant homolog. For example, SF3B1 on chromosome 2, encoding a core spliceosome component, was validated as a CYCLOPS gene in cancers with partial copy loss. Experimental suppression of SF3B1 caused synthetic lethality in these cells, underscoring single-copy gene dependencies as a key therapeutic target (Paolella et al. 2017). Another example is PSMC2, a proteasome subunit encoded on chromosome 7 whose deletion renders cancer cells highly susceptible to further proteasome inhibition. PSMC2 is involved in the degradation of misfolded proteins, a process that is already burdened in aneuploid cancer cells due to their increased proteotoxic stress. Targeting proteasome function in PSMC2-deleted cancers leads to severe cytotoxicity, highlighting its potential as a therapeutic target (Nijhawan et al. 2012).

GEMINI: dependencies caused by loss of heterozygosity of gene variants

Loss of heterozygosity (LOH) in essential genes presents an untapped vulnerability in cancer. Many tumors undergo LOH in hundreds of nondriver genes, generating unique differences between cancer and normal cells. The concept of GEMINI (gene expression modulated by copy number imbalance) genes arises from these alterations, where monoallelic inactivation of the retained allele in tumors can selectively eliminate cancer cells while sparing normal tissue (Nichols et al. 2020). A systematic analysis identified 1278 essential genes undergoing LOH in cancer, with a potential for allele-specific targeting via gene editing, RNAi, or small molecules. Experimental validation demonstrated that inactivating PRIM1 or EXOSC8 selectively impaired tumor cell growth in cells harboring LOH in these genes, whereas cells with the nontargeted allele remained viable. Unlike CYCLOPS genes, which require copy loss, GEMINI vulnerabilities include genes that undergo both copy-neutral LOH (CN-LOH) and copy-loss LOH, broadening their therapeutic scope. Furthermore, GEMINI alterations may coexist with CYCLOPS dependencies, creating expanded therapeutic windows. The study highlights the potential for developing allele-specific gene editing and small molecule approaches to exploit GEMINI vulnerabilities in cancer therapy.

Sensitivity of tumors with chromosome Y loss due to DDX3X inhibition

A special case of aneuploidy causing collateral lethality due to loss of paralog gene function is the loss of the Y (LOY) chromosome, a common event in various cancer types in male patients. Chromosomes X and Y contain several paralog genes that exert essential functions in cells: LOY leads to the (complete) loss of Y-linked genes, leaving the cells dependent on the X-linked genes. One of these Y-linked genes is DDX3Y, which encodes an RNA helicase involved in RNA processing and translation. As a consequence, cells with Y chromosome loss become increasingly reliant on DDX3X, the homologous RNA helicase encoded on the X chromosome. Inhibition of DDX3X disrupts RNA metabolism, impairing tumor cell viability while sparing normal cells that retain both helicases. This presents DDX3X as a promising synthetic-lethal target in tumors harboring Y chromosome loss (Qi et al. 2023).

Chromosome 1q amplification and UCK2 dependency

Amplification of chromosome arm 1q is frequently observed in cancers including melanoma, stomach cancer, and ovarian cancer. Cells with chromosome 1q gain are more sensitive to nucleotide analogs (e.g., 3-deazauridine) due to the upregulation of UCK2 (uridine–cytidine kinase 2), a key enzyme in the pyrimidine salvage pathway, encoded on 1q. The increased expression of UCK2 in cells with 1q gain enhances the phosphorylation and activation of nucleoside analog prodrugs, making these cells more susceptible to their cytotoxic effects. In fact, the elevated UCK2 levels in 1q-amplified cells accelerate their conversion into toxic metabolites. This heightened dependency on UCK2-mediated activation creates a specific vulnerability that can be exploited therapeutically by using UCK2-dependent nucleotide analogs to selectively target 1q-amplified tumors while sparing normal cells with lower UCK2 expression (Girish et al. 2023).

Together, these findings highlight the growing potential of targeting aneuploidy-associated vulnerabilities as a strategy for cancer therapy. Both general and chromosome-specific features of aneuploidy create distinct cellular dependencies, ranging from heightened sensitivity to mitotic stress, impaired protein homeostasis, and metabolic imbalances to collateral vulnerabilities arising from gene dosage alterations. Although several of these strategies have shown compelling mechanistic selectivity and preclinical efficacy, their clinical translation remains limited by toxicity, lack of predictive biomarkers, and insufficient patient stratification. Future efforts to refine these approaches through functional genomics, biomarker discovery, and rational combination therapies may ultimately enable the safe and effective targeting of aneuploid cancers in the clinic.

Conclusions and key outstanding questions

Although ubiquitous, the role of aneuploidy in cancer has been a difficult phenomenon to elucidate due to its context dependence, which we are just now beginning to understand and be able to predict with the use of tumor genomic data sets. Novel CRISPR tools make the targeted engineering of aneuploidy possible, greatly improving the ability to generate isogenic models of CSA. The greatest challenges to identifying and studying CSA are the vast number of genetic contexts that could influence the effect of CSA on tumor growth and our ability to generate representative mouse models.

Nevertheless, the currently available strategies have enabled the causal connection of aneuploidy with certain malignant phenotypes including proliferation, immune evasion, and metastasis. Although CIN provides karyotypic heterogeneity that can enable the survival of tumor cells in challenging environments, it can also be pushed to toxic levels as a strategy to selectively target cancer cells. Additionally, aneuploidy itself causes imbalances and resultant stresses and dependencies that present unique opportunities for therapeutic targeting. As more studies uncover direct relationships between aneuploidy and cancer phenotypes, our ability to treat tumors containing these recurrent genetic alterations will grow. Key outstanding questions in the field that will need to be fully addressed in the near future include the following:

1. Is the effect of CIN and general aneuploidy tissue-specific? There is a need for cross-tumor type, meta-analysis, and genomic studies to understand how tissue and genetic contexts influence the consequences of chromosomal instability (CIN) and aneuploidy.

2. Beyond p53, what mechanisms enable human tumors to tolerate aneuploidy? Identifying additional tolerance pathways could reveal critical cellular adaptations and potential therapeutic targets.

3. What are the differential effects of CIN versus stable aneuploidy on cell-extrinsic phenotypes such as immune evasion, metastasis, and interactions with the microenvironment?

4. How can we distinguish between driver and tolerated (or possibly nonexistent “passenger”) aneuploidies? Functional experiments involving the introduction or removal of specific aneuploidies are needed to probe associated stress responses and tumorigenic potential.

5. How does aneuploidy promote tumorigenesis, drive heterogeneity, and contribute to drug resistance, and how can these vulnerabilities be therapeutically exploited or monitored in patients? This encompasses understanding cancer-specific mechanisms, leveraging aneuploidy-related weaknesses, improving detection, and integrating insights from the immune microenvironment.

Addressing these questions will allow us to gain a deeper understanding of aneuploidy and CIN in cancer. This knowledge may lead to effective strategies for harnessing this cancer hallmark for patient stratification and therapy.

Competing interest statement

T.D. is a cofounder of KaryoVerse. The other authors declare no competing interests.