Development of Genome-Derived Tumor Type Prediction to Inform Clinical Cancer Care | Oncology | JAMA Oncology | JAMA Network
[Skip to Navigation]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 18.206.177.17. Please contact the publisher to request reinstatement.
1.
Hyman  DM, Puzanov  I, Subbiah  V,  et al.  Vemurafenib in Multiple Nonmelanoma Cancers with BRAF V600 mutations.  N Engl J Med. 2015;373(8):726-736. doi:10.1056/NEJMoa1502309PubMedGoogle ScholarCrossref
2.
Varghese  AM, Arora  A, Capanu  M,  et al.  Clinical and molecular characterization of patients with cancer of unknown primary in the modern era.  Ann Oncol. 2017;28(12):3015-3021. doi:10.1093/annonc/mdx545PubMedGoogle ScholarCrossref
3.
Golub  TR, Slonim  DK, Tamayo  P,  et al.  Molecular classification of cancer: class discovery and class prediction by gene expression monitoring.  Science. 1999;286(5439):531-537. doi:10.1126/science.286.5439.531PubMedGoogle ScholarCrossref
4.
Greco  FA, Spigel  DR, Yardley  DA, Erlander  MG, Ma  XJ, Hainsworth  JD.  Molecular profiling in unknown primary cancer: accuracy of tissue of origin prediction.  Oncologist. 2010;15(5):500-506. doi:10.1634/theoncologist.2009-0328PubMedGoogle ScholarCrossref
5.
Marquard  AM, Birkbak  NJ, Thomas  CE,  et al.  TumorTracer: a method to identify the tissue of origin from the somatic mutations of a tumor specimen.  BMC Med Genomics. 2015;8:58. doi:10.1186/s12920-015-0130-0PubMedGoogle ScholarCrossref
6.
Moran  S, Martínez-Cardús  A, Sayols  S,  et al.  Epigenetic profiling to classify cancer of unknown primary: a multicentre, retrospective analysis.  Lancet Oncol. 2016;17(10):1386-1395. doi:10.1016/S1470-2045(16)30297-2PubMedGoogle ScholarCrossref
7.
Soh  KP, Szczurek  E, Sakoparnig  T, Beerenwinkel  N.  Predicting cancer type from tumour DNA signatures.  Genome Med. 2017;9(1):104. doi:10.1186/s13073-017-0493-2PubMedGoogle ScholarCrossref
8.
Ferracin  M, Pedriali  M, Veronese  A,  et al.  MicroRNA profiling for the identification of cancers with unknown primary tissue-of-origin.  J Pathol. 2011;225(1):43-53. doi:10.1002/path.2915PubMedGoogle ScholarCrossref
9.
Kang  S, Li  Q, Chen  Q,  et al.  CancerLocator: non-invasive cancer diagnosis and tissue-of-origin prediction using methylation profiles of cell-free DNA.  Genome Biol. 2017;18(1):53. doi:10.1186/s13059-017-1191-5PubMedGoogle ScholarCrossref
10.
Hao  X, Luo  H, Krawczyk  M,  et al.  DNA methylation markers for diagnosis and prognosis of common cancers.  Proc Natl Acad Sci U S A. 2017;114(28):7414-7419. doi:10.1073/pnas.1703577114PubMedGoogle ScholarCrossref
11.
Snyder  MW, Kircher  M, Hill  AJ, Daza  RM, Shendure  J.  Cell-free DNA comprises an in vivo nucleosome footprint that informs its tissues-of-origin.  Cell. 2016;164(1-2):57-68. doi:10.1016/j.cell.2015.11.050PubMedGoogle ScholarCrossref
12.
Chapman  JS, Asthana  S, Cade  L,  et al.  Clinical sequencing contributes to a BRCA-associated cancer rediagnosis that guides an effective therapeutic course.  J Natl Compr Canc Netw. 2015;13(7):835-845. doi:10.6004/jnccn.2015.0101PubMedGoogle ScholarCrossref
13.
Zehir  A, Benayed  R, Shah  RH,  et al.  Mutational landscape of metastatic cancer revealed from prospective clinical sequencing of 10,000 patients.  Nat Med. 2017;23(6):703-713. doi:10.1038/nm.4333PubMedGoogle ScholarCrossref
14.
Cheng  DT, Mitchell  TN, Zehir  A,  et al.  Memorial Sloan Kettering–Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT): a hybridization capture-based next-generation sequencing clinical assay for solid tumor molecular oncology.  J Mol Diagn. 2015;17(3):251-264. doi:10.1016/j.jmoldx.2014.12.006PubMedGoogle ScholarCrossref
15.
Breiman  L.  Random forests.  Mach Learn. 2001;45:5-32. doi:10.1023/A:1010933404324Google ScholarCrossref
16.
Bella  A, Ferri  C, Hernández-Orallo  J, Ramírez-Quintana  MJ.  On the effect of calibration in classifier combination.  Appl Intell. 2013;38(4):566-585. doi:10.1007/s10489-012-0388-2Google ScholarCrossref
17.
Chang  MT, Asthana  S, Gao  SP,  et al.  Identifying recurrent mutations in cancer reveals widespread lineage diversity and mutational specificity.  Nat Biotechnol. 2016;34(2):155-163. doi:10.1038/nbt.3391PubMedGoogle ScholarCrossref
18.
Adalsteinsson  VA, Ha  G, Freeman  SS,  et al.  Scalable whole-exome sequencing of cell-free DNA reveals high concordance with metastatic tumors.  Nat Commun. 2017;8(1):1324. doi:10.1038/s41467-017-00965-yPubMedGoogle ScholarCrossref
19.
Flaherty  KT, Infante  JR, Daud  A,  et al.  Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations.  N Engl J Med. 2012;367(18):1694-1703. doi:10.1056/NEJMoa1210093PubMedGoogle ScholarCrossref
20.
Long  GV, Stroyakovskiy  D, Gogas  H,  et al.  Combined BRAF and MEK inhibition versus BRAF inhibition alone in melanoma.  N Engl J Med. 2014;371(20):1877-1888. doi:10.1056/NEJMoa1406037PubMedGoogle ScholarCrossref
21.
Pentsova  EI, Shah  RH, Tang  J,  et al.  evaluating cancer of the central nervous system through next-generation sequencing of cerebrospinal fluid.  J Clin Oncol. 2016;34(20):2404-2415. doi:10.1200/JCO.2016.66.6487PubMedGoogle ScholarCrossref
22.
Al-Ahmadie  HA, Iyer  G, Lee  BH,  et al.  Frequent somatic CDH1 loss-of-function mutations in plasmacytoid variant bladder cancer.  Nat Genet. 2016;48(4):356-358. doi:10.1038/ng.3503PubMedGoogle ScholarCrossref
23.
Beltran  H, Eng  K, Mosquera  JM,  et al.  Whole-exome sequencing of metastatic cancer and biomarkers of treatment response.  JAMA Oncol. 2015;1(4):466-474. doi:10.1001/jamaoncol.2015.1313PubMedGoogle ScholarCrossref
24.
Sholl  LM, Do  K, Shivdasani  P,  et al.  Institutional implementation of clinical tumor profiling on an unselected cancer population.  JCI Insight. 2016;1(19):e87062. doi:10.1172/jci.insight.87062PubMedGoogle Scholar
25.
Hirshfield  KM, Tolkunov  D, Zhong  H,  et al.  Clinical actionability of comprehensive genomic profiling for management of rare or refractory cancers.  Oncologist. 2016;21(11):1315-1325. doi:10.1634/theoncologist.2016-0049PubMedGoogle ScholarCrossref
26.
Frampton  GM, Fichtenholtz  A, Otto  GA,  et al.  Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing.  Nat Biotechnol. 2013;31(11):1023-1031. doi:10.1038/nbt.2696PubMedGoogle ScholarCrossref
27.
Roychowdhury  S, Iyer  MK, Robinson  DR,  et al.  Personalized oncology through integrative high-throughput sequencing: a pilot study.  Sci Transl Med. 2011;3(111):111ra121. doi:10.1126/scitranslmed.3003161PubMedGoogle Scholar
28.
Singh  RR, Patel  KP, Routbort  MJ,  et al.  Clinical validation of a next-generation sequencing screen for mutational hotspots in 46 cancer-related genes.  J Mol Diagn. 2013;15(5):607-622. doi:10.1016/j.jmoldx.2013.05.003PubMedGoogle ScholarCrossref
29.
Topol  EJ.  High-performance medicine: the convergence of human and artificial intelligence.  Nat Med. 2019;25(1):44-56. doi:10.1038/s41591-018-0300-7PubMedGoogle ScholarCrossref
30.
Shortliffe  EH, Sepúlveda  MJ.  Clinical decision support in the era of artificial intelligence.  JAMA. 2018;320(21):2199-2200. doi:10.1001/jama.2018.17163PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    November 14, 2019

    Development of Genome-Derived Tumor Type Prediction to Inform Clinical Cancer Care

    Author Affiliations
    • 1Human Oncology and Pathogenesis Program, Memorial Sloan Kettering Cancer Center, New York, New York
    • 2Marie-Josée and Henry R. Kravis Center for Molecular Oncology, Memorial Sloan Kettering Cancer Center, New York, New York
    • 3Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, New York
    • 4Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, New York
    • 5Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, New York
    • 6Clinical Research Administration, Memorial Sloan Kettering Cancer Center, New York, New York
    • 7Weill Cornell Medical College, Department of Medicine, Cornell University, New York, New York
    • 8Weill Cornell Medical College, Department of Pathology and Laboratory Medicine, Cornell University, New York, New York
    JAMA Oncol. 2020;6(1):84-91. doi:10.1001/jamaoncol.2019.3985
    Key Points

    Question  To what extent can genomic features revealed by clinical targeted tumor sequencing enable diagnostic accuracy of tumor type?

    Findings  This cohort study used machine learning techniques to construct and train an algorithmic classifier on a cohort of 7791 prospectively sequenced tumors representing 22 cancer types to predict cancer type and origin from DNA sequence data obtained at the point of care. In some cases, genome-directed reassessment of diagnosis prompted tumor type reclassification resulting in altered therapy for patients with cancer.

    Meaning  The clinical implementation of artificial intelligence to guide tumor type diagnosis at the point of care may complement standard histopathologic testing and imaging to enable improved diagnostic accuracy.

    Abstract

    Importance  Diagnosing the site of origin for cancer is a pillar of disease classification that has directed clinical care for more than a century. Even in an era of precision oncologic practice, in which treatment is increasingly informed by the presence or absence of mutant genes responsible for cancer growth and progression, tumor origin remains a critical factor in tumor biologic characteristics and therapeutic sensitivity.

    Objective  To evaluate whether data derived from routine clinical DNA sequencing of tumors could complement conventional approaches to enable improved diagnostic accuracy.

    Design, Setting, and Participants  A machine learning approach was developed to predict tumor type from targeted panel DNA sequence data obtained at the point of care, incorporating both discrete molecular alterations and inferred features such as mutational signatures. This algorithm was trained on 7791 tumors representing 22 cancer types selected from a prospectively sequenced cohort of patients with advanced cancer.

    Results  The correct tumor type was predicted for 5748 of the 7791 patients (73.8%) in the training set as well as 8623 of 11 644 patients (74.1%) in an independent cohort. Predictions were assigned probabilities that reflected empirical accuracy, with 3388 cases (43.5%) representing high-confidence predictions (>95% probability). Informative molecular features and feature categories varied widely by tumor type. Genomic analysis of plasma cell-free DNA yielded accurate predictions in 45 of 60 cases (75.0%), suggesting that this approach may be applied in diverse clinical settings including as an adjunct to cancer screening. Likely tissues of origin were predicted from targeted tumor sequencing in 95 of 141 patients (67.4%) with cancers of unknown primary site. Applying this method prospectively to patients under active care enabled genome-directed reassessment of diagnosis in 2 patients initially presumed to have metastatic breast cancer, leading to the selection of more appropriate treatments, which elicited clinical responses.

    Conclusions and Relevance  These results suggest that the application of artificial intelligence to predict tissue of origin in oncologic practice can act as a useful complement to conventional histologic review to provide integrated pathologic diagnoses, often with important therapeutic implications.

    ×