Aneuploidy is near-ubiquitous in cancer and contributes to tumor biology. However, the temporal evolutionary dynamics that select for aneuploidy remain uncharacterized. We performed longitudinal genomic analysis of 755 samples from 167 patients with colorectal-derived neoplasias from different stages through metastasis and treatment. Adenomas had few copy number alterations (CNA) and most were subclonal, whereas cancers had many clonal CNAs, suggesting that progression goes through a CNA bottleneck. Individual colorectal cancer glands from the same tumor had similar karyotypes, despite evidence of ongoing instability at the cell level. CNAs in metastatic lesions, after therapy, and in late recurrences were similar to the primary. Mathematical modeling indicated that these data are consistent with the action of negative selection on CNAs that "trap" cancer genomes on a fitness peak characterized by specific CNAs. Hence, progression to colorectal cancer requires traversing a rugged fitness landscape, whereas subsequent CNA evolution is constrained by negative selection. SIGNIFICANCE:We profiled 167 long-term responders longitudinally (755 samples), documenting long-term cancer evolution. We found that a genetic bottleneck is required for progression and is associated with dramatic increase in CNAs but decrease in clonal diversity. After initiation, copy number evolution is constrained by negative selection through metastasis and treatment. See related commentary by Okada et al., p. 192.

Mean fluorescence of OVCAR4 (GFP positive) and Ov4Cis (RFP positive) for up to 20 passages in vitro (mean±st.d, n=3 technical replicates).

Drug resistance results in poor outcomes for patients with cancer. Adaptive therapy is a potential strategy to address drug resistance that exploits competitive interactions between sensitive and resistant subclones. In this study, we showed that adapting carboplatin dose according to tumor response (adaptive therapy) significantly prolonged survival of murine ovarian cancer models compared with standard carboplatin dosing, without increasing mean daily drug dose or toxicity. Platinum-resistant ovarian cancer cells exhibited diminished fitness when drug was absent in vitro and in vivo, which caused selective decline of resistant populations due to reduced proliferation and increased apoptosis. Conversely, fitter, sensitive cells regrew when drug was withdrawn. Using a bioinformatics pipeline that exploits copy number changes to quantify the emergence of treatment resistance, analysis of cell-free DNA obtained longitudinally from patients with ovarian cancer during treatment showed subclonal selection through therapy, and measurements of resistant population growth correlated strongly with disease burden. These preclinical findings pave the way for future clinical testing of personalized adaptive therapy regimens tailored to the evolution of carboplatin resistance in individual patients with ovarian cancer. SIGNIFICANCE:Carboplatin adaptive therapy improves treatment efficacy without increasing daily dose due to reduced fitness of drug-resistant populations, which can be tracked using cfDNA and could direct adaptive therapy in future clinical trials. See related commentary by Gatenby, p. 3373.

Epigenetic alterations co-evolve with genetic mutations to drive carcinogenesis and treatment response. Resolving genome-epigenome coevolution requires accurate multi-omic single cell measurement. Here we develop a new technology called “double ATAC” (dATAC) for high-throughput, high-quality, concurrent whole genome sequencing and chromatin accessibility profiling of individual somatic cells. dATAC is a “one pot” method that uses two rounds of tagmentation to sequentially label open chromatin regions and then the whole genome, and produces data of the same quality as current leading single-omic methods. Using colorectal cancer as a model system, we apply dATAC to reveal convergent reorganisation of the epigenome across expanding drug-resistant clones during 5-FU chemotherapy exposure, remarkable stoichiometry of chromatin accessibility at somatic copy number alterations, and the clonal expansion of copy-number altered T cells in the stroma of metastatic disease. dATAC is a robust single cell technology to accurately profile genome-epigenome coevolution across tissues, diseases and species. ### Competing Interest Statement TG and AMB are coinventors on a patent application that describe a method for TCR sequencing (GB2305655.9). TG is a coinventor on a patent application describing a method to measure evolutionary dynamics in cancers using DNA methylation (GB2317139.0). TG and FW are coinventors on a patent application describing a method to infer drug resistance mechanisms from barcoding data (GB2501439.0). TG has received honorarium from Genentech and consultancy fees from DAiNA therapeutics.

LiqCNA estimates of the emergent resistant population correlate with disease burder. A–C, LiqCNA estimation of resistant disease (percentage, red; right y‐axis) and CA125 (kIU/L, blue; left y‐axis) plotted over time for five unique patients. Time 0 = diagnosis. Subclonal ratio of tumor and blood samples is shown by triangles and squares, respectively. Dotted line indicates the start of each new line of chemotherapy, P, platinum‐containing chemotherapy; dashed line, date of patient death. D, Rate of change in the subclonal ratio estimated by LiqCNA in each sequential time point (compared with the diagnostic sample for that patient) plotted against CA125 at the later time point.

Cancer treatment frequently fails due to the evolution of drug-resistant cell phenotypes driven by genetic or non-genetic changes. The origin, timing, and rate of spread of these adaptations are critical for understanding drug resistance mechanisms but remain challenging to observe directly. We present a mathematical framework to infer drug resistance dynamics from genetic lineage tracing and population size data without direct measurement of resistance phenotypes. Simulation experiments demonstrate that the framework accurately recovers ground-truth evolutionary dynamics. Experimental evolution to 5-Fu chemotherapy in colorectal cancer cell lines SW620 and HCT116 validates the framework. In SW620 cells, a stable pre-existing resistant subpopulation was inferred, whereas in HCT116 cells, resistance emerged through phenotypic switching into a slow-growing resistant state with stochastic progression to full resistance. Functional assays, including scRNA-seq and scDNA-seq, validate these distinct evolutionary routes. This framework facilitates rapid characterisation of resistance mechanisms across diverse experimental settings.

Cumulative dose of carboplatin in mg/kg over time for all mice receiving adaptive therapy. Green=100% sensitive, blue=80% sensitive, red=100% resistant with a different shade for each mouse. A-C: Cumulative carboplatin (mg) plotted against volume of individual tumours separated into A: OVCAR4, sensitive, green, B: 80:20 OVCAR4:Ov4Carbo-Luc, blue and C: Ov4Carbo-Luc, resistant, red. Each line indicates one tumour and the same colour shade is used for the same mouse in Fig.2C. D: Carboplatin dose per day according to injected cell ratio and treatment group. mean±st.d, n=2‐5 mice per group. *p<0.05, paired t‐test.

Cells were seeded in 10% FBS-containing media and after 24 hours, media was changed to either 10% or 0.5% FBS-containing media. Media was exchanged for new media with the same FBS concentration every 24 hours. Cells were imaged every four hours for 120 hours with a 10x incucyte brightfield microscope to measure % confluence. mean±s.d., N=4 biological repeat experiments.

BACKGROUND:The risk of developing advanced neoplasia (AN; colorectal cancer and/or high-grade dysplasia) in ulcerative colitis (UC) patients with a low-grade dysplasia (LGD) lesion is variable and difficult to predict. This is a major challenge for effective clinical management. OBJECTIVE:We aimed to provide accurate AN risk stratification in UC patients with LGD. We hypothesised that the pattern and burden of somatic genomic copy number alterations (CNAs) in LGD lesions could predict future AN risk. DESIGN:We performed a retrospective multicentre validated case-control study using n=270 LGD samples from n=122 patients with UC. Patients were designated progressors (n=40) if they had a diagnosis of AN in the ~5 years following LGD diagnosis or non-progressors (n=82) if they remained AN-free during follow-up. DNA was extracted from the baseline LGD lesion, low-coverage whole genome sequencing performed and data processed to detect CNAs. Survival analysis was used to evaluate CNAs as predictors of future AN risk. RESULTS:CNA burden was significantly higher in progressors than non-progressors (p=2×10-6 in discovery cohort) and was a very significant predictor of AN risk in univariate analysis (OR=36; p=9×10-7), outperforming existing clinical risk factors such as lesion size, shape and focality. Optimal risk prediction was achieved with a multivariate model combining CNA burden with the known clinical risk factor of incomplete LGD resection. Within-LGD lesion genetic heterogeneity did not confound risk prediction. CONCLUSION:Measurement of CNAs in LGD is an accurate predictor of AN risk in inflammatory bowel disease and is likely to support clinical management.

Drug-resistant HGSC exhibits reduced proliferative fitness in low-resource conditions. A, Abundance of sensitive (OVCAR4) and resistant (Ov4Cis) cells cocultured over time in high-resource conditions (10% FBS) at a range of starting ratios and passaged every 3 days. Green, OVCAR4; red, Ov4Cis. Data are shown as mean ± SD, with n = 3 technical replicates. B, Ratio of OVCAR4:OV4Cis in A plotted on a log scale over time with start times staggered (so day 0 of the 85:15 ratio experiment is plotted at the time point when there were 85% sensitive cells remaining in the 95:5 starting ratio experiment). The black line shows a linear fit of this log ratio based on the 5:95 dataset (red stars). The slope of the line corresponds to the difference in growth rate between sensitive and resistant cells: gs − gr = g. C, Sensitive OVCAR4 cells (green; left) and resistant Ov4Cis (red; right) cells were grown as monocultures in 0.5% FBS-containing media that were either not changed (A) or exchanged daily with fresh 0.5% FBS-containing media (B). Mean cell abundance is shown over time. D, Abundance of sensitive (OVCAR4) and resistant (Ov4Cis) cells over time when cocultured without passage in low-resource conditions (0.5% FBS exchanged daily) at three starting ratios (85% sensitive, 50% sensitive, and 15% sensitive). Solid lines indicate measured abundance of sensitive (green) and resistant (red) populations over time. Dashed lines indicate predicted cell growth based on their initial seeding density and their measured growth as monocultures. Data are shown as mean ± SD. Model is based on three technical repeat experiments. E, Ratio of OVCAR4:OV4Cis in D plotted on a log scale over time. The black line shows a linear fit of this log ratio based on the 85:15 dataset.

A: Doxorubicin-induced expression of p16 and p21 in OVCAR4 and Ov4Cis cells by reverse transcription qPCR (mean±st.d, n=3, **p<0.01, unpaired t-test). B: OVCAR4 and Ov4Cis cells were grown as mono-culture and as 85:15 S:R and 50:50 S:R co-cultures. Co-culture samples were sorted into GFP-positive and GFP-negative populations by flow cytometry. Expression of p16 and C: p21 by reverse transcription qPCR in each population is shown (mean±st.d, n=3). Horizontal dotted lines indicate a 2-fold and 0.5-fold increase in gene expression.

Fitness costs of resistance are observed in vivo. A, Schematic depicting in vivo coculture experiments. B, qPCR of GFP and RFP DNA in subcutaneous coculture tumors (OVCAR4:Ov4Cis, 0:100, 10:90, 50:50, 80:20, and 100:0). Tumors were harvested after 12 weeks and qPCR for GFP and RFP is plotted against a standard curve of input DNA (see Supplementary Fig. S7). Solid line, mean; dashed line, SD, one icon per tumor; and n = 2 to 4 tumors per starting ratio. C, IHC for p53, GFP, and cleaved caspase‐3 in 50:50 OVCAR4:Ov4Cis tumors harvested from different mice at weeks 4 and 12. Representative slides are shown at low (top) and high (bottom) magnification. Arrows indicate p53+, GFP-negative resistant cells. Circles, cleaved caspase-3–positive resistant cells; squares, cleaved caspase-3–negative resistant cells. D, Pixel quantification of sensitive and resistant cells. Data are shown as mean ± SD, with n = 2 mice per time point. Solid line indicates the starting ratio of S:R cells.

Locally advanced esophageal adenocarcinoma remains difficult to treat and the ecological and evolutionary dynamics responsible for resistance and recurrence are incompletely understood. Here, we performed longitudinal multiomic analysis of patients with esophageal adenocarcinoma in the MEMORI trial. Multi-region multi-timepoint whole-exome and paired transcriptome sequencing was performed on 27 patients before, during and after neoadjuvant treatment. We found major transcriptomic changes during treatment with upregulation of immune, stromal and oncogenic pathways. Genetic data revealed that clonal sweeps through treatment were rare. Imaging mass cytometry and T cell receptor sequencing revealed remodeling of the tumor microenvironment during treatment. The presence of genetic immune escape, a less-cytotoxic T cell phenotype and a lack of clonal T cell expansions were linked to poor treatment response. In summary, there were widespread transcriptional and environmental changes through treatment, with limited clonal replacement, suggestive of phenotypic plasticity.

Immune system control is a principal hurdle in cancer evolution. The temporal dynamics of immune evasion remain incompletely characterized, and how immune-mediated selection interrelates with epigenome alteration is unclear. Here we infer the genome- and epigenome-driven evolutionary dynamics of tumor-immune coevolution within primary colorectal cancers (CRCs). We utilize a multiregion multiomic dataset of matched genome, transcriptome and chromatin accessibility profiling from 495 single glands (from 29 CRCs) supplemented with high-resolution spatially resolved neoantigen sequencing data and multiplexed imaging of the tumor microenvironment from 82 microbiopsies within 11 CRCs. Somatic chromatin accessibility alterations contribute to accessibility loss of antigen-presenting genes and silencing of neoantigens. Immune escape and exclusion occur at the outset of CRC formation, and later intratumoral differences in immuno-editing are negligible or exclusive to sites of invasion. Collectively, immune evasion in CRC follows a 'Big Bang' evolutionary pattern, whereby it is acquired close to transformation and defines subsequent cancer-immune evolution.

Glioblastoma remains incurable and recurs in all patients. Here we design and characterize a novel induced-recurrence model in which mice xenografted with primary patient-derived glioma initiating/stem cells (GIC) are treated with a therapeutic regimen closely recapitulating patient standard of care, followed by monitoring until tumours recur (induced recurrence patient-derived xenografts, IR-PDX). By tracking in vivo tumour growth, we confirm the patient specificity and initial efficacy of treatment prior to recurrence. Availability of longitudinally matched pairs of primary and recurrent GIC enabled patient-specific evaluation of the fidelity with which the model recapitulated phenotypes associated with the true recurrence. Through comprehensive multi-omic analyses, we show that the IR-PDX model recapitulates aspects of genomic, epigenetic, and transcriptional state heterogeneity upon recurrence in a patient-specific manner. The accuracy of the IR-PDX enabled both novel biological insights, including the positive association between glioblastoma recurrence and levels of ciliated neural stem cell-like tumour cells, and the identification of druggable patient-specific therapeutic vulnerabilities. This proof-of-concept study opens the possibility for prospective precision medicine approaches to identify target-drug candidates for treatment at glioblastoma recurrence.

A: Animal weight over time for mice enrolled in the experiment shown in Fig.4B, using the same colour codes. *= AT-treated OVCAR4 mouse that was culled before experimental end point due to unexplained weight loss with no tumour seen at necropsy. B: Mean animal weight over time according to injected cell type: green=OVCAR4, blue=80:20 OVCAR4:Ov4Carbo, red=Ov4Carbo-Luc, and treatment group: triangle = vehicle, circle=ST and square=AT. n=2-5 per group.

Sensitive and resistant HGSC populations grow and decline dynamically during treatment and can be tracked in cfDNA from patients with HGSC. A, OVCAR4 and Ov4Cis cells were grown as 50:50 cocultures in low-resource conditions. On day 6, cells were treated with cisplatin (0.1–1 μmol/L) or vehicle (dashed line). Cultures were harvested over time and the ratio of sensitive (green) to resistant (red) cells was measured by flow cytometry over time. Data are shown as mean ± SD, with n = 3 technical replicates. B, Tumor purity–corrected copy number profile for each sample obtained from Patient 1. Gray bars show the copy number (CN) profile of diagnostic samples, and red bars show CN profile of later samples as indicated. Transparent bars show genome segments with no resistance‐specific CNAs (i.e., regions with clonally shared CN or with other CNAs). Resistant proportion of each sample estimated by LiqCNA is shown. Error bars indicate the 95% confidence interval of each estimate. C, Zoomed‐in profiles of selected chromosomes with the most prominent/impactful resistance‐specific CNAs. Driver genes or genes associated with ovarian cancer that overlap with resistance‐specific CNA are indicated by blue vertical lines and listed below each graph.

Drug resistance results in dismal outcomes for cancer patients. Carboplatin is the backbone of treatment for high grade serous ovarian cancer (HGSOC), augmented by PARP inhibitor (PARPi) maintenance therapy for patients with tumors deficient in homologous recombination repair (HRR). However, most tumors evolve resistance to both drugs and no other agents, including immunotherapies, have improved survival. Therapeutic strategies that acknowledge and target this evolution offer an opportunity to preserve drug sensitivity and extend progression-free survival. Adaptive Therapy (AT) is one such approach. AT is based on the premise that fitness costs are incurred as resistance evolves, so that competition between drug-sensitive and resistant cells favors sensitive cells when drug is absent. We have developed AT in HGSOC from proof-of-concept to a randomized phase 2 trial called ACTOv, (Adaptive ChemoTherapy in Ovarian cancer) comparing carboplatin AT to standard carboplatin dosing in patients with relapsed, platinum sensitive high grade serous and endometrioid ovarian cancer. We are now extending our work to explore the evolutionary dynamics between PARPi-sensitive and PARPi-resistant HGSOC populations. We evolved resistance to cisplatin and carboplatin in HGSOC cell lines in vitro and in vivo via carboplatin treatment of mice with intraperitoneal xenografts. Resistance to cisplatin, carboplatin and the PARPis niraparib and olaparib was quantified by live/dead assays in the entire cell panel. Co-cultures of sensitive and resistant cells were created in vitro in low resource conditions (low serum or low glucose) to expose fitness deficits. Population growth dynamics were quantified over time using the Sartorius Incucyte™ Live Cell Analysis System. Mechanisms underpinning these dynamics were explored by flow cytometry to analyze cell cycle and expression of the standard apoptotic maker, Annexin V. HRR status was determined functionally by immunofluorescence for Rad51 and γH2Ax co-localization and for micronuclei formation, while BRCA expression was examined by western blot. In vivo co-cultures of GFP-expressing sensitive cells and RFP-expressing resistant cells enabled tracking of sensitive/resistant cell growth over time by qPCR. Sensitive/resistant cells were further examined by immunohistochemistry (IHC) in excised tumors. The influence of drug therapy on in vitro co-cultures of sensitive (GFP) and resistant (RFP) HGSOC was quantified by flow cytometry. AT was compared to standard carboplatin dosing in mice with HGSOC xenografts and excised tumors were examined by IHC. Cumulative drug dose and murine survival were compared. Cell-free DNA (cfDNA) and tumor biopsies were obtained from HGSOC patients during carboplatin and PARPi treatment and subjected to low pass whole genome sequencing and analysis with our published bioinformatic pipeline (LiquidCNA) that estimates the size of the emergent resistant population. Finally, established patient datasets were examined to adapt our pre-clinical discoveries for clinical testing in the ACTOv trial. Cell viability assays confirmed platinum resistance. One murine HGSOC cell line with evolved platinum resistance (60577R3) was also resistant to olaparib and niraparib in vitro, whereas there was no difference in PARPi sensitivity in the other sensitive/resistant cell pairs. Fitness deficits of resistance in the absence of drug, were exposed by low resource co-cultures in vitro in four out of five HGSOC cell lines with evolved platinum resistance. Conversely, the resistant subline, 60577R3, exhibited greater fitness than sensitive, ancestral cells (60577). Resistant populations with reduced fitness declined due to both reduced proliferation and increased apoptosis in vitro and in vivo when drug was absent. Rad51 foci formation in response to double strand breaks was observed in platinum and PARPi-resistant 60577R3 cells but not sensitive 60577 indicating that functional HRR had been restored in 60577R3 as resistance evolved. Consistent with this, 60577R3 cells also made fewer segregation errors during mitosis. Western blot confirmed expression of BRCA1 in 60577R3 but not 60577 cells. Platinum treatment of in vitro co-cultures demonstrated that the size of sensitive and resistant HGSOC populations fluctuated with drug therapy and that fitter, sensitive cells re-grew when drug was withdrawn. In vivo, AT significantly prolonged survival of murine ovarian cancer models compared to standard carboplatin dosing without increasing mean daily drug dose or drug-related toxicity. Escape from control by AT was associated with growth of resistant cells in our in vivo models. In HGSOC patients, LiqCNA analysis of cfDNA obtained longitudinally during treatment showed sub-clonal selection through therapy and demonstrated that LiqCNA correlates with increases in the serum tumor marker, CA125, a proxy for disease progression. Existing clinical datasets of CA125 during standard carboplatin treatment defined the parameters for dose adaptations in patients receiving AT in the ACTOv trial. Platinum resistance can be associated with fitness deficits in HGSOC, which can be exploited by AT to preserve platinum sensitivity over the long-term and prolong survival. BRCA reversion mutations are the most well-documented cause of acquired PARPi resistance. Our data imply that this may provide a fitness advantage for HGSOC cells with implications for PARPi maintenance treatment and response to subsequent lines of therapy. These discoveries paved the way for us to launch the ACTOv trial, which is currently recruiting patients in ten NHS hospitals throughout the UK. ACTOv patients will provide biopsies and very frequent blood samples for circulating tumor DNA (total ∼1,400 samples from 80 patients), enabling us to interrogate clonal evolution during standard and adaptive therapy and relate this to clinical outcome. LiqCNA could provide a biomarker for the emergence of drug resistance to direct AT in second generation clinical trials. Mohammed Ateeb Khan, Helen Hockings, Maximilian Mossner, Ann-Marie Baker, Hall Amy, Robert L. Hollis, Weini Huang, Eszter Lakatos, Charlie Gourley, Trevor A. Graham, Michelle Lockley. Translating adaptive therapy to ovarian cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr SY43-01.