Prediction of plasma ctDNA fraction and prognostic implications of liquid biopsy in advanced prostate cancer.

No consensus strategies exist for prognosticating metastatic castration-resistant prostate cancer (mCRPC). Circulating tumor DNA fraction (ctDNA%) is increasingly reported by commercial and laboratory tests but its utility for risk stratification is unclear. Here, we intersect ctDNA%, treatment outcomes, and clinical characteristics across 738 plasma samples from 491 male mCRPC patients from two randomized multicentre phase II trials and a prospective province-wide blood biobanking program. ctDNA% correlates with serum and radiographic metrics of disease burden and is highest in patients with liver metastases. ctDNA% strongly predicts overall survival, progression-free survival, and treatment response independent of therapeutic context and outperformed established prognostic clinical factors. Recognizing that ctDNA-based biomarker genotyping is limited by low ctDNA% in some patients, we leverage the relationship between clinical prognostic factors and ctDNA% to develop a clinically-interpretable machine-learning tool that predicts whether a patient has sufficient ctDNA% for informative ctDNA genotyping (available online: https://www.ctDNA.org ). Our results affirm ctDNA% as an actionable tool for patient risk stratification and provide a practical framework for optimized biomarker testing.

Multiregion sampling of de novo metastatic prostate cancer reveals complex polyclonality and augments clinical genotyping.

De novo metastatic prostate cancer is highly aggressive, but the paucity of routinely collected tissue has hindered genomic stratification and precision oncology. Here, we leveraged a rare study of surgical intervention in 43 de novo metastatic prostate cancers to assess somatic genotypes across 607 synchronous primary and metastatic tissue regions plus circulating tumor DNA. Intra-prostate heterogeneity was pervasive and impacted clinically relevant genes, resulting in discordant genotypes between select primary restricted regions and synchronous metastases. Additional complexity was driven by polyclonal metastatic seeding from phylogenetically related primary populations. When simulating clinical practice relying on a single tissue region, genomic heterogeneity plus variable tumor fraction across samples caused inaccurate genotyping of dominant disease; however, pooling extracted DNA from multiple biopsy cores before sequencing can rescue misassigned somatic genotypes. Our results define the relationship between synchronous treatment-sensitive primary and metastatic lesions in men with de novo metastatic prostate cancer and provide a framework for implementing genomics-guided patient management.

A generalizable machine learning framework for classifying DNA repair defects using ctDNA exomes.

Specific classes of DNA damage repair (DDR) defect can drive sensitivity to emerging therapies for metastatic prostate cancer. However, biomarker approaches based on DDR gene sequencing do not accurately predict DDR deficiency or treatment benefit. Somatic alteration signatures may identify DDR deficiency but historically require whole-genome sequencing of tumour tissue. We assembled whole-exome sequencing data for 155 high ctDNA fraction plasma cell-free DNA and matched leukocyte DNA samples from patients with metastatic prostate or bladder cancer. Labels for DDR gene alterations were established using deep targeted sequencing. Per sample mutation and copy number features were used to train XGBoost ensemble models. Naive somatic features and trinucleotide signatures were associated with specific DDR gene alterations but insufficient to resolve each class. Conversely, XGBoost-derived models showed strong performance including an area under the curve of 0.99, 0.99 and 1.00 for identifying BRCA2, CDK12, and mismatch repair deficiency in metastatic prostate cancer. Our machine learning approach re-classified several samples exhibiting genomic features inconsistent with original labels, identified a metastatic bladder cancer sample with a homozygous BRCA2 copy loss, and outperformed an existing exome-based classifier for BRCA2 deficiency. We present DARC Sign (DnA Repair Classification SIGNatures); a public machine learning tool leveraging clinically-practical liquid biopsy specimens for simultaneously identifying multiple types of metastatic prostate cancer DDR deficiencies. We posit that it will be useful for understanding differential responses to DDR-directed therapies in ongoing clinical trials and may ultimately enable prospective identification of prostate cancers with phenotypic evidence of DDR deficiency.

Deep whole-genome ctDNA chronology of treatment-resistant prostate cancer.

Circulating tumour DNA (ctDNA) in blood plasma is an emerging tool for clinical cancer genotyping and longitudinal disease monitoring1. However, owing to past emphasis on targeted and low-resolution profiling approaches, our understanding of the distinct populations that comprise bulk ctDNA is incomplete2-12. Here we perform deep whole-genome sequencing of serial plasma and synchronous metastases in patients with aggressive prostate cancer. We comprehensively assess all classes of genomic alterations and show that ctDNA contains multiple dominant populations, the evolutionary histories of which frequently indicate whole-genome doubling and shifts in mutational processes. Although tissue and ctDNA showed concordant clonally expanded cancer driver alterations, most individual metastases contributed only a minor share of total ctDNA. By comparing serial ctDNA before and after clinical progression on potent inhibitors of the androgen receptor (AR) pathway, we reveal population restructuring converging solely on AR augmentation as the dominant genomic driver of acquired treatment resistance. Finally, we leverage nucleosome footprints in ctDNA to infer mRNA expression in synchronously biopsied metastases, including treatment-induced changes in AR transcription factor signalling activity. Our results provide insights into cancer biology and show that liquid biopsy can be used as a tool for comprehensive multi-omic discovery.

Genomic Features of Lung-Recurrent Hormone-Sensitive Prostate Cancer.

PURPOSE: Pulmonary involvement is rare in metastatic hormone-sensitive prostate cancer (mHSPC) that recurs after treatment for localized disease. Guidelines recommend intensive systemic therapy, similar to patients with liver metastases, but some lung-recurrent mHSPC may have good outcomes. Genomic features of lung metastases may clarify disease aggression, but are poorly understood since lung biopsy is rarely performed. We present a comparative assessment of genomic drivers and heterogeneity in metachronous prostate tumors and lung metastases. METHODS: We leveraged a prospective functional imaging study of 208 biochemically recurrent prostate cancers to identify 10 patients with lung-recurrent mHSPC. Histologic diagnosis was attained via thoracic surgery or fine-needle lung biopsy. We retrieved clinical data and performed multiregion sampling of primary tumors and metastases. Targeted and/or whole-exome sequencing was applied to 46 primary and 32 metastatic foci. RESULTS: Unusually for mHSPC, all patients remained alive despite a median follow-up of 11.5 years. Several patients experienced long-term freedom from systemic treatment. The genomic landscape of lung-recurrent mHSPC was typical of curable prostate cancer with frequent PTEN, SPOP, and chromosome 8p alterations, and there were no deleterious TP53 and DNA damage repair gene mutations that characterize aggressive prostate cancer. Despite a long median time to recurrence (76.8 months), copy number alterations and clonal mutations were highly conserved between metastatic and primary foci, consistent with intrapatient homogeneity and limited genomic evolution. CONCLUSION: In this retrospective hypothesis-generating study, we observed indolent genomic etiology in selected lung-recurrent mHSPC, cautioning against grouping these patients together with liver or bone-predominant mHSPC. Although our data do not generalize to all patients with lung metastases, the results encourage prospective efforts to stratify lung-recurrent mHSPC by genomic features.

Genetic analysis of nucleotide-binding leucine-rich repeat (NLR) receptors in multiple sclerosis.

Genetic and functional analyses of the inflammasome suggest a role for this multiprotein complex in the biological mechanisms leading to the onset and progression of multiple sclerosis (MS). Nucleotide-binding, leucine-rich repeat (NLR) receptors trigger the activation and assembly of specific inflammasomes in response to danger signals. Mining exome sequencing data from 326 MS patients identified 17 rare missense or nonsense variants in NLR family pyrin domain containing 1 (NLRP1), NLRP3, NLRP6, NLRP7 and NLR family CARD domain containing 4 (NLRC4). Genotyping these variants in 2503 MS cases and 1076 healthy controls did not result in statistically significant differences between groups, and segregation analysis within MS families was largely unsupportive of co-segregation of these variants with disease. However, the identification of MS patients harboring rare homozygote variants in NLRP1 (p.Ile601Phe and p.Ser1387Ile), a variant in NLRP3 (p.Leu832Ile) resulting in the substitution of a critical amino acid for the formation of its leucine-rich repeat domain, and several MS patients with NLRC4 variants (p.Arg310Ter and p.Glu600Ter) causing protein truncations suggest that rare protein-altering variants in inflammasome-activating NLR receptors may contribute to MS risk.

Exome sequencing in multiple sclerosis families identifies 12 candidate genes and nominates biological pathways for the genesis of disease.

Multiple sclerosis (MS) is an inflammatory disease of the central nervous system characterized by myelin loss and neuronal dysfunction. Although the majority of patients do not present familial aggregation, Mendelian forms have been described. We performed whole-exome sequencing analysis in 132 patients from 34 multi-incident families, which nominated likely pathogenic variants for MS in 12 genes of the innate immune system that regulate the transcription and activation of inflammatory mediators. Rare missense or nonsense variants were identified in genes of the fibrinolysis and complement pathways (PLAU, MASP1, C2), inflammasome assembly (NLRP12), Wnt signaling (UBR2, CTNNA3, NFATC2, RNF213), nuclear receptor complexes (NCOA3), and cation channels and exchangers (KCNG4, SLC24A6, SLC8B1). These genes suggest a disruption of interconnected immunological and pro-inflammatory pathways as the initial event in the pathophysiology of familial MS, and provide the molecular and biological rationale for the chronic inflammation, demyelination and neurodegeneration observed in MS patients.

Whole-Exome Sequencing of an Exceptional Longevity Cohort.

Centenarians represent a unique cohort to study the genetic basis for longevity and factors determining the risk of neurodegenerative disorders, including Alzheimer's disease (AD). The estimated genetic contribution to longevity is highest in centenarians and super-cententenarians, but few genetic variants have been shown to clearly impact this phenotype. While the genetic risk for AD and other dementias is now well understood, the frequency of known dementia risk variants in centenarians is not fully characterized. To address these questions, we performed whole-exome sequencing on 100 individuals of 98-108 years age in search of genes with large effect sizes towards the exceptional aging phenotype. Overall, we were unable to identify a rare protein-altering variant or individual genes with an increased burden of rare variants associated with exceptional longevity. Gene burden analysis revealed three genes of nominal statistical significance associated with extreme aging, including LYST, MDN1, and RBMXL1. Several genes with variants conferring an increased risk for AD and other dementias were identified, including TREM2, EPHA1, ABCA7, PLD3, MAPT, and NOTCH3. Larger centenarian studies will be required to further elucidate the genetic basis for longevity, and factors conferring protection against age-dependent neurodegenerative syndromes.

Analysis of NOD-like receptor NLRP1 in multiple sclerosis families.

The implementation of exome sequencing technologies has started to unravel the genetic etiology of familial multiple sclerosis (MS). A homozygote p.G587S mutation in NLRP1 has been suggested as potentially causative for the onset of MS in an affected sibling pair, who later developed malignant melanoma. To validate the proposed role of recessive NLRP1 mutations in the pathological mechanisms of MS, we examined exome sequencing data from 326 MS patients from Canada for the identification of NLRP1 missense and nonsense variants. This analysis did not identify the previously described p.G587S mutation; however, three patients with potential NLRP1 compound heterozygote mutations were observed. Haplotype and segregation analyses indicate that the variants observed in these patients were inherited in cis, and do not segregate with disease within families. Thus, the analysis of MS patients from Canada failed to identify potentially pathogenic mutations in NLRP1, including the previously described p.G587S mutation. Further studies are necessary to confirm a role of NLRP1 in the pathophysiology of MS.

Genetic modifiers of multiple sclerosis progression, severity and onset.

The genetic contribution to clinical outcomes for multiple sclerosis (MS) has yet to be defined. We performed exome sequencing analysis in 100 MS patients presenting opposite extremes of clinical phenotype (discovery cohort), and genotyped variants of interest in 2016 MS patients (replication cohort). Linear and logistic regression analyses were used to identify significant associations with disease course, severity and onset. Our analysis of the discovery cohort nominated 38 variants in 21 genes. Replication analysis identified PSMG4 p.W99R and NLRP5 p.M459I to be associated with disease severity (p=0.002 and 0.008). CACNA1H p.R1871Q was found associated with patients presenting relapsing remitting MS at clinical onset (p=0.028) whereas NLRP5 p.M459I and EIF2AK1 p.K558R were associated with primary progressive disease (p=0.031 and 0.023). In addition, PSMG4 p.W99R and NLRP5 p.R761L were found to correlate with an earlier age at MS clinical onset, and MC1R p.R160W with delayed onset of clinical symptoms (p=0.010-0.041).

Purinergic receptors P2RX4 and P2RX7 in familial multiple sclerosis.

Genetic variants in the purinergic receptors P2RX4 and P2RX7 have been shown to affect susceptibility to multiple sclerosis (MS). In this study, we set out to evaluate whether rare coding variants of major effect could also be identified in these purinergic receptors. Sequencing analysis of P2RX4 and P2RX7 in 193 MS patients and 100 controls led to the identification of a rare three variant haplotype (P2RX7 rs140915863:C>T [p.T205M], P2RX7 rs201921967:A>G [p.N361S], and P2RX4 rs765866317:G>A [p.G135S]) segregating with disease in a multi-incident family with six family members diagnosed with MS (logarithm of odds = 3.07). Functional analysis of this haplotype in HEK293 cells revealed impaired P2X7 surface expression (P < 0.01), resulting in over 95% inhibition of adenosine triphosphate (ATP)-induced pore function (P < 0.001) and a marked reduction in phagocytic ability (P < 0.05). In addition, transfected cells showed 40% increased peak ATP-induced inward current (P < 0.01), and a greater Ca2+ response to the P2X4 135S variant compared with wild type (P < 0.0001). Our study nominates rare genetic variants in P2RX4 and P2RX7 as major genetic contributors to disease, further supporting a role for these purinergic receptors in MS and the disruption of transmembrane cation channels leading to impairment of phagocytosis as the pathological mechanisms of disease.

Common genetic etiology between "multiple sclerosis-like" single-gene disorders and familial multiple sclerosis.

Several single-gene disorders with clinical and radiological characteristics similar to those observed in multiple sclerosis (MS) patients have been described. To evaluate whether this phenotypic overlap can be ascribed to a common genetic etiology, 28 genes known to present pathogenic mutations for 24 of these disorders were sequenced in 270 MS patients. All identified variants were genotyped in 2131 MS cases and 830 healthy controls, and those exclusively observed in patients were assessed for segregation within families. This analysis identified 9 rare variants in 6 genes segregating with disease in 13 families. Four different mutations were identified in CYP27A1, including a reported pathogenic mutation for cerebrotendinous xanthomatosis (p.R405W), which was observed in six patients from a multi-incident family, three diagnosed with MS, two with an undefined neurological disease and one seemingly healthy. A LYST p.V1678A and a PDHA1 p.K387Q mutation were both observed in five MS patients from three separate multi-incident families. In addition, CLCN2 p.V174G, GALC p.D162E and POLG p.R361G were each identified in two MS patients from one family. This study suggests a shared genetic etiology between MS and the characterized single-gene disorders, and highlights cholesterol metabolism and the synthesis of oxysterols as important biological mechanisms for familial MS.

Case-Control Studies Are Not Familial Studies.

Identifying rare genetic variants that drive the onset of disease is challenging, even before considering the additional genetic and environmental influences that likely exist in complex diseases. We recently published a study proposing a rare variant in the NR1H3 gene (p.R415Q, rs61731956) as responsible for the onset of multiple sclerosis (MS) in two multi-incident families (Wang et al., 2016). This publication has generated much discussion, and fortunately the possibility to validate a finding or prove it spurious can occur rapidly in genetic studies. All novel discoveries must be replicated, and best efforts should be made to ensure that these replications use the appropriate samples and approach, and provide the correct interpretation of the results. This Matters Arising Response paper addresses the Minikel and MacArthur (2016) and The International Multiple Sclerosis Genetics Consortium (2016) Matters Arising papers, published concurrently in Neuron.

Nuclear Receptor NR1H3 in Familial Multiple Sclerosis.

Multiple sclerosis (MS) is an inflammatory disease characterized by myelin loss and neuronal dysfunction. Despite the aggregation observed in some families, pathogenic mutations have remained elusive. In this study, we describe the identification of NR1H3 p.Arg415Gln in seven MS patients from two multi-incident families presenting severe and progressive disease, with an average age at onset of 34 years. Additionally, association analysis of common variants in NR1H3 identified rs2279238 conferring a 1.35-fold increased risk of developing progressive MS. The p.Arg415Gln position is highly conserved in orthologs and paralogs, and disrupts NR1H3 heterodimerization and transcriptional activation of target genes. Protein expression analysis revealed that mutant NR1H3 (LXRA) alters gene expression profiles, suggesting a disruption in transcriptional regulation as one of the mechanisms underlying MS pathogenesis. Our study indicates that pharmacological activation of LXRA or its targets may lead to effective treatments for the highly debilitating and currently untreatable progressive phase of MS.

Analysis of Plasminogen Genetic Variants in Multiple Sclerosis Patients.

Multiple sclerosis (MS) is a prevalent neurological disease of complex etiology. Here, we describe the characterization of a multi-incident MS family that nominated a rare missense variant (p.G420D) in plasminogen (PLG) as a putative genetic risk factor for MS. Genotyping of PLG p.G420D (rs139071351) in 2160 MS patients, and 886 controls from Canada, identified 10 additional probands, two sporadic patients and one control with the variant. Segregation in families harboring the rs139071351 variant, identified p.G420D in 26 out of 30 family members diagnosed with MS, 14 unaffected parents, and 12 out of 30 family members not diagnosed with disease. Despite considerably reduced penetrance, linkage analysis supports cosegregation of PLG p.G420D and disease. Genotyping of PLG p.G420D in 14446 patients, and 8797 controls from Canada, France, Spain, Germany, Belgium, and Austria failed to identify significant association with disease (P = 0.117), despite an overall higher prevalence in patients (OR = 1.32; 95% CI = 0.93-1.87). To assess whether additional rare variants have an effect on MS risk, we sequenced PLG in 293 probands, and genotyped all rare variants in cases and controls. This analysis identified nine rare missense variants, and although three of them were exclusively observed in MS patients, segregation does not support pathogenicity. PLG is a plausible biological candidate for MS owing to its involvement in immune system response, blood-brain barrier permeability, and myelin degradation. Moreover, components of its activation cascade have been shown to present increased activity or expression in MS patients compared to controls; further studies are needed to clarify whether PLG is involved in MS susceptibility.

Analysis of CH25H in multiple sclerosis and neuromyelitis optica.

Oxysterols produced by CH25H during cholesterol metabolism have been shown to play an important role in the immune response. In this study we report the genetic characterization of CH25H in patients diagnosed with multiples sclerosis (MS) or neuromyelitis optica (NMO), for the identification of genetic variants affecting disease susceptibility and course. Exome analysis in 212 MS and 14 NMO patients identified a rare CH25H p.Q17H mutation in two NMO patients of Asian ancestry. In addition, association analysis of common CH25H variants identified a nominally significant association with MS patients presenting a clinical course consistent with primary progressive disease.