Phylogenetic reconstruction of adult blood cancer reveals early origins and lifelong evolution. | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Biological Sciences - Article Phylogenetic reconstruction of adult blood cancer reveals early origins and lifelong evolution. Jyoti Nangalia, Nicholas Williams, Joe Lee, Luiza Moore, E Baxter, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-93830/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 20 Jan, 2022 Read the published version in Nature → Version 1 posted You are reading this latest preprint version Abstract Mutations in cancer-associated genes drive tumour outgrowth. However, the timing of driver mutations and dynamics of clonal expansion that lead to human cancers are largely unknown. We used 448,553 somatic mutations from whole-genome sequencing of 843 clonal haematopoietic colonies to reconstruct the phylogeny of haematopoiesis, from embryogenesis to clinical disease, in 10 patients with myeloproliferative neoplasms which are blood cancers more common in older age. JAK2V617F, the pathognomonic mutation in these cancers, was acquired in utero or childhood, with upper estimates of age of acquisition ranging between 4.1 months and 11.4 years across 5 patients. DNMT3A mutations, which are associated with age-related clonal haematopoiesis, were also acquired in utero or childhood, by 7.9 weeks of gestation to 7.8 years across 4 patients. Subsequent driver mutation acquisition was separated by decades. The mean latency between JAK2V617F acquisition and clinical presentation was 34 years (range 20-54 years). Rates of clonal expansion varied substantially (200% expansion/year), were affected by additional driver mutations, and predicted latency to clinical presentation. Driver mutations and rates of expansion would have been detectable in blood one to four decades before clinical presentation. This study reveals how driver mutation acquisition very early in life with life-long growth trajectories drive adult blood cancer, providing opportunities for early detection and intervention, and a new paradigm for cancer development. Cancer Biology Oncology phylogenetic reconstruction blood cancer mutations Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Full Text Due to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the latest manuscript can be downloaded and accessed as a PDF. Additional Declarations There is NO Competing Interest. Cite Share Download PDF Status: Published Journal Publication published 20 Jan, 2022 Read the published version in Nature → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-93830","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Biological Sciences - Article","associatedPublications":[],"authors":[{"id":3768170,"identity":"dc0b599e-05cb-4e30-be76-aadcbe5eb486","order_by":0,"name":"Jyoti Nangalia","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwklEQVRIiWNgGAWjYBACxgb+hw+QBQyI0MLDjKKKsBYGBh42CdK0MM/uPVb5o+aOXb/0AcYPPxgOGxN22Jxzabd5jj1LntmXwCzZw3DYjLCWGQlmtxnYDicbnGFgkGZgOGxDlJbCH/8OJ9ufYWD+TaSWHDMG3rbDdgY8DGwgW4hxWFqyNG/f4QSJM4xtlj0G6YS9bzgj+eDHH98O2/P3MB++8aPC2rCBoBaoisQGYKwSF5HyUNqeCLWjYBSMglEwUgEAvHs5allex54AAAAASUVORK5CYII=","orcid":"","institution":"Wellcome Sanger Institute","correspondingAuthor":true,"prefix":"","firstName":"Jyoti","middleName":"","lastName":"Nangalia","suffix":""},{"id":3768171,"identity":"ee6ed54b-a847-4070-89c1-1e3b6d94b521","order_by":1,"name":"Nicholas Williams","email":"","orcid":"https://orcid.org/0000-0003-3989-9167","institution":"Wellcome Trust Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Nicholas","middleName":"","lastName":"Williams","suffix":""},{"id":3768172,"identity":"c01a2783-2e22-4f4d-847a-a4d0c2a5b9e9","order_by":2,"name":"Joe Lee","email":"","orcid":"","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Joe","middleName":"","lastName":"Lee","suffix":""},{"id":3768173,"identity":"8f0f01e9-a834-4707-b011-ccb8e2fcd250","order_by":3,"name":"Luiza Moore","email":"","orcid":"https://orcid.org/0000-0001-5315-516X","institution":"Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Luiza","middleName":"","lastName":"Moore","suffix":""},{"id":3768174,"identity":"715d4ee8-4f51-4ce5-946f-0fcd9eccf7e0","order_by":4,"name":"E Baxter","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"E","middleName":"","lastName":"Baxter","suffix":""},{"id":3768175,"identity":"c6b602c1-85de-4028-9663-4aac3b8f13ca","order_by":5,"name":"James Hewinson","email":"","orcid":"","institution":"Wellcome Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"","lastName":"Hewinson","suffix":""},{"id":3768176,"identity":"7fc2340c-90f5-45db-a291-f46521c68312","order_by":6,"name":"Kevin Dawson","email":"","orcid":"","institution":"The Cancer, Ageing and Somatic Mutation Programme, Wellcome Trust Sanger Institute, Hinxton, Cambridgeshire CB10 1SA","correspondingAuthor":false,"prefix":"","firstName":"Kevin","middleName":"","lastName":"Dawson","suffix":""},{"id":3768177,"identity":"7dce4d6b-0d85-4d90-9e19-bfebf84d03c4","order_by":7,"name":"Andrew Menzies","email":"","orcid":"","institution":"Wellcome Trust Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"","lastName":"Menzies","suffix":""},{"id":3768178,"identity":"09420082-98d6-4bd6-b1e3-c932591982b9","order_by":8,"name":"Anna Godfrey","email":"","orcid":"","institution":"Cambridge Universities NHS Trust","correspondingAuthor":false,"prefix":"","firstName":"Anna","middleName":"","lastName":"Godfrey","suffix":""},{"id":3768179,"identity":"58fcdc7f-e7fd-464e-aaa7-c51ab483f977","order_by":9,"name":"Anthony Green","email":"","orcid":"","institution":"University of Cambridge","correspondingAuthor":false,"prefix":"","firstName":"Anthony","middleName":"","lastName":"Green","suffix":""},{"id":3768180,"identity":"e80ae891-424a-4e78-8267-12503ada4513","order_by":10,"name":"Peter Campbell","email":"","orcid":"https://orcid.org/0000-0002-3921-0510","institution":"The Cancer Ageing and Somatic Mutation Programme Wellcome Trust Sanger Institute","correspondingAuthor":false,"prefix":"","firstName":"Peter","middleName":"","lastName":"Campbell","suffix":""}],"badges":[],"createdAt":"2020-10-16 15:41:55","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-93830/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-93830/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41586-021-04312-6","type":"published","date":"2022-01-20T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":3546146,"identity":"4b346dcb-e3b6-4272-b584-22008fbf7515","added_by":"auto","created_at":"2020-11-12 19:36:50","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":172559,"visible":true,"origin":"","legend":"Patient cohort and experimental design\nA. Experimental design. WGS, whole genome sequencing; BFU-E, Burst forming unit-erythroid. B. Patient cohort\nshowing ages at diagnosis, disease phase and duration of disease, sample types and timepoints. ET, Essential\nthrombocythemia; PV polycythemia vera; MF, myelofibrosis. The length of the shaded bars represents the\nduration of disease, either to last follow-up or to patient death. C. Driver mutations, both single nucleotide\nvariants and insertions/deletions, as well as copy number aberrations identified in at least one colony within\neach patient are shown. Shaded colours represent the type of mutation and the numbers within the squares\nrepresent the number of mutations or copy number aberrations in individual patients.","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1/bcef6ffd17c8ff9da4e266ce.png"},{"id":3546147,"identity":"58828fdb-fcde-44bc-ba69-65fdd9625944","added_by":"auto","created_at":"2020-11-12 19:36:50","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":136643,"visible":true,"origin":"","legend":"Phylogenetic histories of 3 patients with MPN driven by JAK2V617F\nThe phylogenetic trees for 3 patients with stable JAK2V617F-mutated MPN diagnosed at different ages. PD7271,\na 21 years old female, presented with asymptomatic isolated thrombocytosis in keeping with ET, and was\ntreated with aspirin. PD5163, a 32 years old female, presented with splanchnic vein thrombosis, relatively\nnormal blood count parameters, a raised red cell mass in keeping with PV, and was treated with Interferonalpha.\nPD5117 was diagnosed with asymptomatic PV at age 64 on the basis of elevated blood counts and a red\ncell mass, and was treated with hydroxycarbamide. The tips of the branches represent individual colonies (red\ndots). Shared branches represent those mutations present across all downstream descendant colonies, and an\nend branch represents mutations unique to the single colony at its branch tip. Branch lengths are proportional\nto mutation counts shown on the vertical axes. Branches containing driver mutations and chromosomal\naberrations are highlighted on the trees by colour. The corresponding times for the start and end of the shared\nbranches harbouring driver mutations are shown on the trees. Ages at diagnosis and any progression of disease\nare shown in labelling above each tree, and ages at the time of sampling are shown to the left of trees. For\nbranches with copy number aberrations, such as chromosome 9p uniparental disomy (UPD) in the phylogenetic\ntree of PD5117, we show the B-allele frequency (BAF) plots of part of chromosome 9p to highlight the\nchromosome breakpoints (vertical red line) for each acquisition. Heterozygous SNVs are mutations that occur\nafter 9pUPD (vertical green line), whereas homozygous SNVs (that are not germline SNVs) would have been\nheterozygous SNVs prior to the 9pUPD but become homozygous as a consequence of the UPD. Given a clonespecific\nmutation rate, the proportion of heterozygous to homozygous SNVs on 9p can broadly indicate the\ntiming of the UPD event. In this case, the leftmost 9pUPD occurred prior to other 9pUPD events due to the\ngreater number of heterozygous mutations that have accumulated since acquisition. ET, Essential\nThrombocythemia; PV, Polycythemia Vera.","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1/540d29b600893ae615b127ec.png"},{"id":3546148,"identity":"1db3b797-3256-400b-b994-95e92c66c626","added_by":"auto","created_at":"2020-11-12 19:36:50","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":216185,"visible":true,"origin":"","legend":"Phylogenetic histories of 7 patients with JAK2V617F-mutated MPN and clonal evolution.\nThe phylogenetic trees of the remaining 7 patients with MPN who have evidence of multiple driver mutation\nled expansions. The vertical axis shows mutation counts. The tips of the branches represent individual colonies.\nSome patients were sampled at multiple timepoints, each timepoint highlighted by different coloured dots at branch ends. Age at diagnosis, times of any disease transformation, driver mutations and timing of mutations\nare depicted. *The timing of 9pUPD events in PD4781 and PD5847 are calculated using the proportion of\nheterozygous versus homozygous mutations on the UPD regions following estimations of clade-specific\nmutation rates. ET, essential thrombocythemia; PV, polycythemia vera; MF, myelofibrosis.","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1/8ea711d531a36f5a82cb799c.png"},{"id":3546149,"identity":"0193f1e9-8989-4bee-874f-9c8273a74b6f","added_by":"auto","created_at":"2020-11-12 19:36:50","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":174119,"visible":true,"origin":"","legend":"Mutation rates and impact of driver mutations\nA. Total single nucleotide variants (SNV) and relationship to age. Dots represent single colonies that underwent\nwhole genome sequencing and colours represent individual patients. Total SNVs represent non-germline SNVs\nadjusted for depth of sequencing. The black line shows the regression line and grey shading shows the 95%\nconfidence interval. B. Clade specific mutation rates across individual patients. Patients and genotypes of clades\nare shown on the left. WT, wildtype clades are shown in grey bars, JAK2-mutated clades are shown in red and\nother mutant clades are shown in yellow. Number of colonies within each clade is shown on the right. The\ncohort wide estimate for the mutation rate in WT colonies is shown by the dotted black vertical line. C.\nRelationship between mean telomere length and age, for wildtype (grey dots), JAK2-mutated (red dots) and\nother mutant colonies (yellow dots). P-values *\u003c0.5, **\u003c0.01, ***\u003c0.001 with multiple hypothesis correction.","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1/a91e8735d9bfe73930961315.png"},{"id":3546150,"identity":"ced17345-eefe-427c-a7ae-9a076ac0c53b","added_by":"auto","created_at":"2020-11-12 19:36:51","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":238029,"visible":true,"origin":"","legend":"Clonal fitness and early detection\nA. We define the fitness of clones by the selection coefficient, S, as the degree of clonal expansion occurring\nevery year. S = 1 implies 100% additional growth, whereas S = 0 implies no change in clone size. The table shows\nS for clades across the cohort, ranked from highest to lowest, along with 95% confidence intervals (CI). S is\nhighest for multiply mutated clades (2.33/year), and lowest for driver mutations common in clonal\nhaematopoiesis (0.09/year). Coloured shading of rows are individual patients harbouring several different\nclades. B. The latency to diagnosis in relation to S following acquisition of mutated-JAK2 is shown for 5 patients\n(PD7271, PD5163, PD5117, PD6646 and PD6629). Red dots represent patients with only mutated-JAK2 as the\ndriver mutation. Black dots represent JAK2 mutation acquisition following mutated-DNMT3A C. The lowest and\nhighest S in the context of the single driver mutation JAK2V617F, demonstrating the changing clonal fractions\nover the life of the patients. D. The modelled relationship between S, final variant allele fraction at MPN\ndiagnosis, and the detection gap in years, assuming assay sensitivities of 0.1%, 1% and 5%. E. The lowest and\nhighest S in the cohort detected in the same individual, from a very slowly growing in utero acquired mutated-\nDNMT3A clone, to a multiply mutated rapidly growing MPN clone. Pink arrowheads show age at diagnosis.","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1/60a8ad352bc03bb706575687.png"},{"id":19033042,"identity":"b900ba03-1b9b-479b-b665-91795ea3caf4","added_by":"auto","created_at":"2022-03-09 17:49:03","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9281575,"visible":true,"origin":"","legend":"","description":"","filename":"20201106Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1_covered.pdf"},{"id":13555237,"identity":"86e28aea-42cc-4daf-bffd-5415d81200b3","added_by":"auto","created_at":"2021-09-17 02:44:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9276549,"visible":true,"origin":"","legend":"","description":"","filename":"20201106Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1_covered.pdf"},{"id":10199749,"identity":"9ea663fc-2d2c-4ac1-86ff-50ea2ce22896","added_by":"auto","created_at":"2021-06-10 10:32:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9272514,"visible":true,"origin":"","legend":"","description":"","filename":"20201106Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1_covered.pdf"},{"id":3546151,"identity":"d23a772b-dfa0-4165-bc06-884bdc0df05d","added_by":"auto","created_at":"2020-11-12 19:37:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":12365636,"visible":true,"origin":"","legend":"","description":"","filename":"20201106Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-93830/v1_stamped.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Phylogenetic reconstruction of adult blood cancer reveals early origins and lifelong evolution.","fulltext":[{"header":"Full Text","content":"Due to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the latest manuscript can be downloaded and \u003ca href='/article/rs-93830/latest.pdf' target='_blank'\u003e accessed as a PDF.\u003c/a\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"
[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"phylogenetic reconstruction, blood cancer, mutations","lastPublishedDoi":"10.21203/rs.3.rs-93830/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-93830/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Mutations in cancer-associated genes drive tumour outgrowth. However, the timing of driver mutations and dynamics of clonal expansion that lead to human cancers are largely unknown. We used 448,553 somatic mutations from whole-genome sequencing of 843 clonal haematopoietic colonies to reconstruct the phylogeny of haematopoiesis, from embryogenesis to clinical disease, in 10 patients with myeloproliferative neoplasms which are blood cancers more common in older age. JAK2V617F, the pathognomonic mutation in these cancers, was acquired in utero or childhood, with upper estimates of age of acquisition ranging between 4.1 months and 11.4 years across 5 patients. DNMT3A mutations, which are associated with age-related clonal haematopoiesis, were also acquired in utero or childhood, by 7.9 weeks of gestation to 7.8 years across 4 patients. Subsequent driver mutation acquisition was separated by decades. The mean latency between JAK2V617F acquisition and clinical presentation was 34 years (range 20-54 years). Rates of clonal expansion varied substantially (\u003c10% to \u003e200% expansion/year), were affected by additional driver mutations, and predicted latency to clinical presentation. Driver mutations and rates of expansion would have been detectable in blood one to four decades before clinical presentation. This study reveals how driver mutation acquisition very early in life with life-long growth trajectories drive adult blood cancer, providing opportunities for early detection and intervention, and a new paradigm for cancer development.","manuscriptTitle":"Phylogenetic reconstruction of adult blood cancer reveals early origins and lifelong evolution.","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2020-11-12 19:36:48","doi":"10.21203/rs.3.rs-93830/v1","editorialEvents":[],"status":"published","journal":{"display":false,"email":"
[email protected]","identity":"nature","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nature","sideBox":"Learn more about [Nature](http://www.nature.com/nature/)","snPcode":"","submissionUrl":"","title":"Nature","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"33f0f272-8d1d-4ce7-9061-9b6fc7cbc4dc","owner":[],"postedDate":"November 12th, 2020","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":1051639,"name":"Cancer Biology"},{"id":1051640,"name":"Oncology"}],"tags":[],"updatedAt":"2022-03-09T17:48:36+00:00","versionOfRecord":{"articleIdentity":"rs-93830","link":"https://doi.org/10.1038/s41586-021-04312-6","journal":{"identity":"nature","isVorOnly":false,"title":"Nature"},"publishedOn":"2022-01-20 05:00:00","publishedOnDateReadable":"January 20th, 2022"},"versionCreatedAt":"2020-11-12 19:36:48","video":"","vorDoi":"10.1038/s41586-021-04312-6","vorDoiUrl":"https://doi.org/10.1038/s41586-021-04312-6","workflowStages":[]},"version":"v1","identity":"rs-93830","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-93830","identity":"rs-93830","version":["v1"]},"buildId":"_2-kVJe1T_tPrBINL-cwx","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.