AI-based automation of enrollment criteria and endpoint assessment in clinical trials in liver diseases.

Clinical trials in metabolic dysfunction-associated steatohepatitis (MASH, formerly known as nonalcoholic steatohepatitis) require histologic scoring for assessment of inclusion criteria and endpoints. However, variability in interpretation has impacted clinical trial outcomes. We developed an artificial intelligence-based measurement (AIM) tool for scoring MASH histology (AIM-MASH). AIM-MASH predictions for MASH Clinical Research Network necroinflammation grades and fibrosis stages were reproducible (κ = 1) and aligned with expert pathologist consensus scores (κ = 0.62-0.74). The AIM-MASH versus consensus agreements were comparable to average pathologists for MASH Clinical Research Network scores (82% versus 81%) and fibrosis (97% versus 96%). Continuous scores produced by AIM-MASH for key histological features of MASH correlated with mean pathologist scores and noninvasive biomarkers and strongly predicted progression-free survival in patients with stage 3 (P < 0.0001) and stage 4 (P = 0.03) fibrosis. In a retrospective analysis of the ATLAS trial (NCT03449446), responders receiving study treatment showed a greater continuous change in fibrosis compared with placebo (P = 0.02). Overall, these results suggest that AIM-MASH may assist pathologists in histologic review of MASH clinical trials, reducing inter-rater variability on trial outcomes and offering a more sensitive and reproducible measure of patient responses.

AI powered quantification of nuclear morphology in cancers enables prediction of genome instability and prognosis.

While alterations in nucleus size, shape, and color are ubiquitous in cancer, comprehensive quantification of nuclear morphology across a whole-slide histologic image remains a challenge. Here, we describe the development of a pan-tissue, deep learning-based digital pathology pipeline for exhaustive nucleus detection, segmentation, and classification and the utility of this pipeline for nuclear morphologic biomarker discovery. Manually-collected nucleus annotations were used to train an object detection and segmentation model for identifying nuclei, which was deployed to segment nuclei in H&E-stained slides from the BRCA, LUAD, and PRAD TCGA cohorts. Interpretable features describing the shape, size, color, and texture of each nucleus were extracted from segmented nuclei and compared to measurements of genomic instability, gene expression, and prognosis. The nuclear segmentation and classification model trained herein performed comparably to previously reported models. Features extracted from the model revealed differences sufficient to distinguish between BRCA, LUAD, and PRAD. Furthermore, cancer cell nuclear area was associated with increased aneuploidy score and homologous recombination deficiency. In BRCA, increased fibroblast nuclear area was indicative of poor progression-free and overall survival and was associated with gene expression signatures related to extracellular matrix remodeling and anti-tumor immunity. Thus, we developed a powerful pan-tissue approach for nucleus segmentation and featurization, enabling the construction of predictive models and the identification of features linking nuclear morphology with clinically-relevant prognostic biomarkers across multiple cancer types.

Artificial Intelligence-Powered Assessment of Pathologic Response to Neoadjuvant Atezolizumab in Patients With NSCLC: Results From the LCMC3 Study.

INTRODUCTION: Pathologic response (PathR) by histopathologic assessment of resected specimens may be an early clinical end point associated with long-term outcomes with neoadjuvant therapy. Digital pathology may improve the efficiency and precision of PathR assessment. LCMC3 (NCT02927301) evaluated neoadjuvant atezolizumab in patients with resectable NSCLC and reported a 20% major PathR rate. METHODS: We determined PathR in primary tumor resection specimens using guidelines-based visual techniques and developed a convolutional neural network model using the same criteria to digitally measure the percent viable tumor on whole-slide images. Concordance was evaluated between visual determination of percent viable tumor (n = 151) performed by one of the 47 local pathologists and three central pathologists. RESULTS: For concordance among visual determination of percent viable tumor, the interclass correlation coefficient was 0.87 (95% confidence interval [CI]: 0.84-0.90). Agreement for visually assessed 10% or less viable tumor (major PathR [MPR]) in the primary tumor was 92.1% (Fleiss kappa = 0.83). Digitally assessed percent viable tumor (n = 136) correlated with visual assessment (Pearson r = 0.73; digital/visual slope = 0.28). Digitally assessed MPR predicted visually assessed MPR with outstanding discrimination (area under receiver operating characteristic curve, 0.98) and was associated with longer disease-free survival (hazard ratio [HR] = 0.30; 95% CI: 0.09-0.97, p = 0.033) and overall survival (HR = 0.14, 95% CI: 0.02-1.06, p = 0.027) versus no MPR. Digitally assessed PathR strongly correlated with visual measurements. CONCLUSIONS: Artificial intelligence-powered digital pathology exhibits promise in assisting pathologic assessments in neoadjuvant NSCLC clinical trials. The development of artificial intelligence-powered approaches in clinical settings may aid pathologists in clinical operations, including routine PathR assessments, and subsequently support improved patient care and long-term outcomes.

AI-based histologic scoring enables automated and reproducible assessment of enrollment criteria and endpoints in NASH clinical trials.

Clinical trials in nonalcoholic steatohepatitis (NASH) require histologic scoring for assessment of inclusion criteria and endpoints. However, guidelines for scoring key features have led to variability in interpretation, impacting clinical trial outcomes. We developed an artificial intelligence (AI)-based measurement (AIM) tool for scoring NASH histology (AIM-NASH). AIM-NASH predictions for NASH Clinical Research Network (CRN) grades of necroinflammation and stages of fibrosis aligned with expert consensus scores and were reproducible. Continuous scores produced by AIM-NASH for key histological features of NASH correlated with mean pathologist scores and with noninvasive biomarkers and strongly predicted patient outcomes. In a retrospective analysis of the ATLAS trial, previously unmet pathological endpoints were met when scored by the AIM-NASH algorithm alone. Overall, these results suggest that AIM-NASH may assist pathologists in histologic review of NASH clinical trials, reducing inter-rater variability on trial outcomes and offering a more sensitive and reproducible measure of patient therapeutic response.

Integration of deep learning-based histopathology and transcriptomics reveals key genes associated with fibrogenesis in patients with advanced NASH.

Nonalcoholic steatohepatitis (NASH) is the most common chronic liver disease globally and a leading cause for liver transplantation in the US. Its pathogenesis remains imprecisely defined. We combined two high-resolution modalities to tissue samples from NASH clinical trials, machine learning (ML)-based quantification of histological features and transcriptomics, to identify genes that are associated with disease progression and clinical events. A histopathology-driven 5-gene expression signature predicted disease progression and clinical events in patients with NASH with F3 (pre-cirrhotic) and F4 (cirrhotic) fibrosis. Notably, the Notch signaling pathway and genes implicated in liver-related diseases were enriched in this expression signature. In a validation cohort where pharmacologic intervention improved disease histology, multiple Notch signaling components were suppressed.

Evolutionary history of transformation from chronic lymphocytic leukemia to Richter syndrome.

Richter syndrome (RS) arising from chronic lymphocytic leukemia (CLL) exemplifies an aggressive malignancy that develops from an indolent neoplasm. To decipher the genetics underlying this transformation, we computationally deconvoluted admixtures of CLL and RS cells from 52 patients with RS, evaluating paired CLL-RS whole-exome sequencing data. We discovered RS-specific somatic driver mutations (including IRF2BP2, SRSF1, B2M, DNMT3A and CCND3), recurrent copy-number alterations beyond del(9p21)(CDKN2A/B), whole-genome duplication and chromothripsis, which were confirmed in 45 independent RS cases and in an external set of RS whole genomes. Through unsupervised clustering, clonally related RS was largely distinct from diffuse large B cell lymphoma. We distinguished pathways that were dysregulated in RS versus CLL, and detected clonal evolution of transformation at single-cell resolution, identifying intermediate cell states. Our study defines distinct molecular subtypes of RS and highlights cell-free DNA analysis as a potential tool for early diagnosis and monitoring.

Molecular map of chronic lymphocytic leukemia and its impact on outcome.

Recent advances in cancer characterization have consistently revealed marked heterogeneity, impeding the completion of integrated molecular and clinical maps for each malignancy. Here, we focus on chronic lymphocytic leukemia (CLL), a B cell neoplasm with variable natural history that is conventionally categorized into two subtypes distinguished by extent of somatic mutations in the heavy-chain variable region of immunoglobulin genes (IGHV). To build the 'CLL map,' we integrated genomic, transcriptomic and epigenomic data from 1,148 patients. We identified 202 candidate genetic drivers of CLL (109 new) and refined the characterization of IGHV subtypes, which revealed distinct genomic landscapes and leukemogenic trajectories. Discovery of new gene expression subtypes further subcategorized this neoplasm and proved to be independent prognostic factors. Clinical outcomes were associated with a combination of genetic, epigenetic and gene expression features, further advancing our prognostic paradigm. Overall, this work reveals fresh insights into CLL oncogenesis and prognostication.

ZBTB33 is mutated in clonal hematopoiesis and myelodysplastic syndromes and impacts RNA splicing.

Clonal hematopoiesis results from somatic mutations in cancer driver genes in hematopoietic stem cells. We sought to identify novel drivers of clonal expansion using an unbiased analysis of sequencing data from 84,683 persons and identified common mutations in the 5-methylcytosine reader, ZBTB33, as well as in YLPM1, SRCAP, and ZNF318. We also identified these mutations at low frequency in myelodysplastic syndrome patients. Zbtb33 edited mouse hematopoietic stem and progenitor cells exhibited a competitive advantage in vivo and increased genome-wide intron retention. ZBTB33 mutations potentially link DNA methylation and RNA splicing, the two most commonly mutated pathways in clonal hematopoiesis and MDS.

Multi-platform profiling characterizes molecular subgroups and resistance networks in chronic lymphocytic leukemia.

Knowledge of the genomic landscape of chronic lymphocytic leukemia (CLL) grows increasingly detailed, providing challenges in contextualizing the accumulated information. To define the underlying networks, we here perform a multi-platform molecular characterization. We identify major subgroups characterized by genomic instability (GI) or activation of epithelial-mesenchymal-transition (EMT)-like programs, which subdivide into non-inflammatory and inflammatory subtypes. GI CLL exhibit disruption of genome integrity, DNA-damage response and are associated with mutagenesis mediated through activation-induced cytidine deaminase or defective mismatch repair. TP53 wild-type and mutated/deleted cases constitute a transcriptionally uniform entity in GI CLL and show similarly poor progression-free survival at relapse. EMT-like CLL exhibit high genomic stability, reduced benefit from the addition of rituximab and EMT-like differentiation is inhibited by induction of DNA damage. This work extends the perspective on CLL biology and risk categories in TP53 wild-type CLL. Furthermore, molecular targets identified within each subgroup provide opportunities for new treatment approaches.

Human-interpretable image features derived from densely mapped cancer pathology slides predict diverse molecular phenotypes.

Computational methods have made substantial progress in improving the accuracy and throughput of pathology workflows for diagnostic, prognostic, and genomic prediction. Still, lack of interpretability remains a significant barrier to clinical integration. We present an approach for predicting clinically-relevant molecular phenotypes from whole-slide histopathology images using human-interpretable image features (HIFs). Our method leverages >1.6 million annotations from board-certified pathologists across >5700 samples to train deep learning models for cell and tissue classification that can exhaustively map whole-slide images at two and four micron-resolution. Cell- and tissue-type model outputs are combined into 607 HIFs that quantify specific and biologically-relevant characteristics across five cancer types. We demonstrate that these HIFs correlate with well-known markers of the tumor microenvironment and can predict diverse molecular signatures (AUROC 0.601-0.864), including expression of four immune checkpoint proteins and homologous recombination deficiency, with performance comparable to 'black-box' methods. Our HIF-based approach provides a comprehensive, quantitative, and interpretable window into the composition and spatial architecture of the tumor microenvironment.

Evaluating the molecular diagnostic yield of joint genotyping-based approach for detecting rare germline pathogenic and putative loss-of-function variants.

PURPOSE: Cohort-based germline variant characterization is the standard approach for pathogenic variant discovery in clinical and research samples. However, the impact of cohort size on the molecular diagnostic yield of joint genotyping is largely unknown. METHODS: Head-to-head comparison of the molecular diagnostic yield of joint genotyping in two cohorts of 239 cancer patients in the absence and then in the presence of 100 additional germline exomes. RESULTS: In 239 testicular cancer patients, 4 (7.4%, 95% confidence interval [CI]: 2.1-17.9) of 54 pathogenic variants in the cancer predisposition and American College of Medical Genetics and Genomics (ACMG) genes were missed by one or both computational runs of joint genotyping. Similarly, 8 (12.1%, 95% CI: 5.4-22.5) of 66 pathogenic variants in these genes were undetected by joint genotyping in another independent cohort of 239 breast cancer patients. An exome-wide analysis of putative loss-of-function (pLOF) variants in the testicular cancer cohort showed that 162 (8.2%, 95% CI: 7.1-9.6) pLOF variants were only detected in one analysis run but not the other, while 433 (22.0%, 95% CI: 20.2-23.9%) pLOF variants were filtered out by both analyses despite having sufficient sequencing coverage. CONCLUSION: Our analysis of the standard germline variant detection method highlighted a substantial impact of concurrently analyzing additional genomic data sets on the ability to detect clinically relevant germline pathogenic variants.

A Machine Learning Approach Enables Quantitative Measurement of Liver Histology and Disease Monitoring in NASH.

BACKGROUND AND AIMS: Manual histological assessment is currently the accepted standard for diagnosing and monitoring disease progression in NASH, but is limited by variability in interpretation and insensitivity to change. Thus, there is a critical need for improved tools to assess liver pathology in order to risk stratify NASH patients and monitor treatment response. APPROACH AND RESULTS: Here, we describe a machine learning (ML)-based approach to liver histology assessment, which accurately characterizes disease severity and heterogeneity, and sensitively quantifies treatment response in NASH. We use samples from three randomized controlled trials to build and then validate deep convolutional neural networks to measure key histological features in NASH, including steatosis, inflammation, hepatocellular ballooning, and fibrosis. The ML-based predictions showed strong correlations with expert pathologists and were prognostic of progression to cirrhosis and liver-related clinical events. We developed a heterogeneity-sensitive metric of fibrosis response, the Deep Learning Treatment Assessment Liver Fibrosis score, which measured antifibrotic treatment effects that went undetected by manual pathological staging and was concordant with histological disease progression. CONCLUSIONS: Our ML method has shown reproducibility and sensitivity and was prognostic for disease progression, demonstrating the power of ML to advance our understanding of disease heterogeneity in NASH, risk stratify affected patients, and facilitate the development of therapies.

Detection of Pathogenic Variants With Germline Genetic Testing Using Deep Learning vs Standard Methods in Patients With Prostate Cancer and Melanoma.

Importance: Less than 10% of patients with cancer have detectable pathogenic germline alterations, which may be partially due to incomplete pathogenic variant detection. Objective: To evaluate if deep learning approaches identify more germline pathogenic variants in patients with cancer. Design, Setting, and Participants: A cross-sectional study of a standard germline detection method and a deep learning method in 2 convenience cohorts with prostate cancer and melanoma enrolled in the US and Europe between 2010 and 2017. The final date of clinical data collection was December 2017. Exposures: Germline variant detection using standard or deep learning methods. Main Outcomes and Measures: The primary outcomes included pathogenic variant detection performance in 118 cancer-predisposition genes estimated as sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The secondary outcomes were pathogenic variant detection performance in 59 genes deemed actionable by the American College of Medical Genetics and Genomics (ACMG) and 5197 clinically relevant mendelian genes. True sensitivity and true specificity could not be calculated due to lack of a criterion reference standard, but were estimated as the proportion of true-positive variants and true-negative variants, respectively, identified by each method in a reference variant set that consisted of all variants judged to be valid from either approach. Results: The prostate cancer cohort included 1072 men (mean [SD] age at diagnosis, 63.7 [7.9] years; 857 [79.9%] with European ancestry) and the melanoma cohort included 1295 patients (mean [SD] age at diagnosis, 59.8 [15.6] years; 488 [37.7%] women; 1060 [81.9%] with European ancestry). The deep learning method identified more patients with pathogenic variants in cancer-predisposition genes than the standard method (prostate cancer: 198 vs 182; melanoma: 93 vs 74); sensitivity (prostate cancer: 94.7% vs 87.1% [difference, 7.6%; 95% CI, 2.2% to 13.1%]; melanoma: 74.4% vs 59.2% [difference, 15.2%; 95% CI, 3.7% to 26.7%]), specificity (prostate cancer: 64.0% vs 36.0% [difference, 28.0%; 95% CI, 1.4% to 54.6%]; melanoma: 63.4% vs 36.6% [difference, 26.8%; 95% CI, 17.6% to 35.9%]), PPV (prostate cancer: 95.7% vs 91.9% [difference, 3.8%; 95% CI, -1.0% to 8.4%]; melanoma: 54.4% vs 35.4% [difference, 19.0%; 95% CI, 9.1% to 28.9%]), and NPV (prostate cancer: 59.3% vs 25.0% [difference, 34.3%; 95% CI, 10.9% to 57.6%]; melanoma: 80.8% vs 60.5% [difference, 20.3%; 95% CI, 10.0% to 30.7%]). For the ACMG genes, the sensitivity of the 2 methods was not significantly different in the prostate cancer cohort (94.9% vs 90.6% [difference, 4.3%; 95% CI, -2.3% to 10.9%]), but the deep learning method had a higher sensitivity in the melanoma cohort (71.6% vs 53.7% [difference, 17.9%; 95% CI, 1.82% to 34.0%]). The deep learning method had higher sensitivity in the mendelian genes (prostate cancer: 99.7% vs 95.1% [difference, 4.6%; 95% CI, 3.0% to 6.3%]; melanoma: 91.7% vs 86.2% [difference, 5.5%; 95% CI, 2.2% to 8.8%]). Conclusions and Relevance: Among a convenience sample of 2 independent cohorts of patients with prostate cancer and melanoma, germline genetic testing using deep learning, compared with the current standard genetic testing method, was associated with higher sensitivity and specificity for detection of pathogenic variants. Further research is needed to understand the relevance of these findings with regard to clinical outcomes.

Integrated molecular drivers coordinate biological and clinical states in melanoma.

We performed harmonized molecular and clinical analysis on 1,048 melanomas and discovered markedly different global genomic properties among subtypes (BRAF, (N)RAS, NF1, triple wild-type (TWT)), subtype-specific preferences for secondary driver genes and active mutational processes previously unreported in melanoma. Secondary driver genes significantly enriched in specific subtypes reflected preferential dysregulation of additional pathways, such as induction of transforming growth factor-ß signaling in BRAF melanomas and inactivation of the SWI/SNF complex in (N)RAS melanomas, and select co-mutation patterns coordinated selective response to immune checkpoint blockade. We also defined the mutational landscape of TWT melanomas and revealed enrichment of DNA-repair-defect signatures in this subtype, which were associated with transcriptional downregulation of key DNA-repair genes, and may revive previously discarded or currently unconsidered therapeutic modalities for genomically stratified melanoma patient subsets. Broadly, harmonized meta-analysis of melanoma whole exomes revealed distinct molecular drivers that may point to multiple opportunities for biological and therapeutic investigation.

Genomic Profiling of Smoldering Multiple Myeloma Identifies Patients at a High Risk of Disease Progression.

PURPOSE: Smoldering multiple myeloma (SMM) is a precursor condition of multiple myeloma (MM) with a 10% annual risk of progression. Various prognostic models exist for risk stratification; however, those are based on solely clinical metrics. The discovery of genomic alterations that underlie disease progression to MM could improve current risk models. METHODS: We used next-generation sequencing to study 214 patients with SMM. We performed whole-exome sequencing on 166 tumors, including 5 with serial samples, and deep targeted sequencing on 48 tumors. RESULTS: We observed that most of the genetic alterations necessary for progression have already been acquired by the diagnosis of SMM. Particularly, we found that alterations of the mitogen-activated protein kinase pathway (KRAS and NRAS single nucleotide variants [SNVs]), the DNA repair pathway (deletion 17p, TP53, and ATM SNVs), and MYC (translocations or copy number variations) were all independent risk factors of progression after accounting for clinical risk staging. We validated these findings in an external SMM cohort by showing that patients who have any of these three features have a higher risk of progressing to MM. Moreover, APOBEC associated mutations were enriched in patients who progressed and were associated with a shorter time to progression in our cohort. CONCLUSION: SMM is a genetically mature entity whereby most driver genetic alterations have already occurred, which suggests the existence of a right-skewed model of genetic evolution from monoclonal gammopathy of undetermined significance to MM. We identified and externally validated genomic predictors of progression that could distinguish patients at high risk of progression to MM and, thus, improve on the precision of current clinical models.

Identification of cancer driver genes based on nucleotide context.

Cancer genomes contain large numbers of somatic mutations but few of these mutations drive tumor development. Current approaches either identify driver genes on the basis of mutational recurrence or approximate the functional consequences of nonsynonymous mutations by using bioinformatic scores. Passenger mutations are enriched in characteristic nucleotide contexts, whereas driver mutations occur in functional positions, which are not necessarily surrounded by a particular nucleotide context. We observed that mutations in contexts that deviate from the characteristic contexts around passenger mutations provide a signal in favor of driver genes. We therefore developed a method that combines this feature with the signals traditionally used for driver-gene identification. We applied our method to whole-exome sequencing data from 11,873 tumor-normal pairs and identified 460 driver genes that clustered into 21 cancer-related pathways. Our study provides a resource of driver genes across 28 tumor types with additional driver genes identified according to mutations in unusual nucleotide contexts.

Scaling computational genomics to millions of individuals with GPUs.

Current genomics methods are designed to handle tens to thousands of samples but will need to scale to millions to match the pace of data and hypothesis generation in biomedical science. Here, we show that high efficiency at low cost can be achieved by leveraging general-purpose libraries for computing using graphics processing units (GPUs), such as PyTorch and TensorFlow. We demonstrate > 200-fold decreases in runtime and ~ 5-10-fold reductions in cost relative to CPUs. We anticipate that the accessibility of these libraries will lead to a widespread adoption of GPUs in computational genomics.

Passenger Hotspot Mutations in Cancer.

Current statistical models for assessing hotspot significance do not properly account for variation in site-specific mutability, thereby yielding many false-positives. We thus (i) detail a Log-normal-Poisson (LNP) background model that accounts for this variability in a manner consistent with models of mutagenesis; (ii) use it to show that passenger hotspots arise from all common mutational processes; and (iii) apply it to a ∼10,000-patient cohort to nominate driver hotspots with far fewer false-positives compared with conventional methods. Overall, we show that many cancer hotspot mutations recurring at the same genomic site across multiple tumors are actually passenger events, recurring at inherently mutable genomic sites under no positive selection.

Genomic Evolutionary Patterns of Leiomyosarcoma and Liposarcoma.

PURPOSE: Leiomyosarcoma and liposarcoma are common subtypes of soft tissue sarcoma (STS). Patients with metastatic leiomyosarcoma or dedifferentiated liposarcoma (DDLPS) typically have worse outcomes compared with localized leiomyosarcoma or well-differentiated liposarcoma (WDLPS). A better understanding of genetic changes between primary/metastatic leiomyosarcoma and between WDLPS/DDLPS may provide insight into their genetic evolution. EXPERIMENTAL DESIGN: We interrogated whole-exome sequencing (WES) from "trios" of normal tissue, primary tumor, and metastatic tumor from individual patients with leiomyosarcoma (n = 9), and trios of normal tissue, well-differentiated tumor, and dedifferentiated tumor from individual patients with liposarcoma (n = 19). Specifically, we performed mutational, copy number, and tumor evolution analyses on these cohorts and compared patterns among leiomyosarcoma and liposarcoma trios. RESULTS: Leiomyosarcoma cases harbored shared drivers through a typical parent/child relationship where the metastatic tumor was derived from the primary tumor. In contrast, while all liposarcoma cases shared the characteristic focal chromosome 12 amplicon, most paired liposarcoma cases did not share additional mutations, suggesting a divergent evolutionary pattern from a common precursor. No highly recurrent genomic alterations from WES were identified that could be implicated as driving the progression of disease in either sarcoma subtype. CONCLUSIONS: From a genomic perspective, leiomyosarcoma metastases contain genetic alterations that are also found in primary tumors. WDLPS and DDLPS, however, appear to divergently evolve from a common precursor harboring 12q amplification, rather than as a transformation to a higher-grade tumor. Further efforts to identify specific drivers of these distinct evolutionary patterns may inform future translational and clinical research in STS.

Growth dynamics in naturally progressing chronic lymphocytic leukaemia.

How the genomic features of a patient's cancer relate to individual disease kinetics remains poorly understood. Here we used the indolent growth dynamics of chronic lymphocytic leukaemia (CLL) to analyse the growth rates and corresponding genomic patterns of leukaemia cells from 107 patients with CLL, spanning decades-long disease courses. We found that CLL commonly demonstrates not only exponential expansion but also logistic growth, which is sigmoidal and reaches a certain steady-state level. Each growth pattern was associated with marked differences in genetic composition, the pace of disease progression and the extent of clonal evolution. In a subset of patients, whose serial samples underwent next-generation sequencing, we found that dynamic changes in the disease course of CLL were shaped by the genetic events that were already present in the early slow-growing stages. Finally, by analysing the growth rates of subclones compared with their parental clones, we quantified the growth advantage conferred by putative CLL drivers in vivo.