Detection and localization of surgically resectable cancers with a multi-analyte blood test

See allHide authors and affiliations

Science  23 Feb 2018:
Vol. 359, Issue 6378, pp. 926-930
DOI: 10.1126/science.aar3247

eLetters is an online forum for ongoing peer review. Submission of eLetters are open to all. eLetters are not edited, proofread, or indexed.  Please read our Terms of Service before submitting your own eLetter.

Compose eLetter

Plain text

  • Plain text
    No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Author Information
First or given name, e.g. 'Peter'.
Your last, or family, name, e.g. 'MacMoody'.
Your email address, e.g.
Your role and/or occupation, e.g. 'Orthopedic Surgeon'.
Your organization or institution (if applicable), e.g. 'Royal Free Hospital'.
Statement of Competing Interests

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.

Enter the characters shown in the image.

Vertical Tabs

  • RE: Detection and localization of surgically resectable cancers with a multi-analyte blood test
    • Chenkai Ma, PhD, Department of Surgery, Royal Melbourne Hospital, University of Melbourne

    Cohen et al proposed a multi-analyte blood test combined with ctDNA and protein for cancer diagnosis in January of Science online where they demonstrated this test (CancerSEEK) could achieve 70% of a median sensitivity and >99% of specificity in eight common types of cancer (Cohen et al., 2018). Their promising results prime the application of ‘liquid biopsy’ in cancer management. However, there are some issues need to be considered before clinical utility of cancer screening.
    The selection of testing genes panels need to be reviewed further. A total of 16 genes involving 61 amplicons were utilized in CancerSEEK for detection of ctDNA from patients’ plasma samples but not all these genes were identified to be associated with the early event of tumorigenesis and other significant to cancer diagnosis genes such as HER2 in non-small cell lung cancer were not included. In addition, TP53 is the most common mutation identified from both cancer patients and healthy controls. How to interpret these mutations in healthy controls remains challenges for physicians. If the proportion of mutant genes fragments in wild types and the total concentration of plasma DNA, a standard analysis pipeline (including robust and producible algorithm) should be established for eliminating the bias between labs.
    The source of ctDNA is mainly derived from necrotic or apoptotic tumour cells, as a result, the diagnostic power of ctDNA might be limited for cancer screening. A previous stud...

    Show More
    Competing Interests: None declared.
  • RE: Detection and localization of surgically resectable cancers with a multi-analyte blood test
    • Michel Wissing, Postdoctoral Fellow, Department of Oncology, McGill University, Montreal, QC, Canada
    • Other Contributors:
      • Eduardo Franco, Professor, Department of Oncology, McGill University, Montreal, QC, Canada

    This article introduced CancerSEEK as a minimally invasive screening test to detect various cancers (1). The authors elegantly combine the detection of genetic mutations in circulating tumor DNA with increased expression of protein biomarkers to optimize the detection of eight cancers in their case population, while limiting false positives in controls. However, challenges and opportunities exist that, if addressed, can maximize the potential of CancerSEEK as a screening tool.

    Considering that carcinogenesis in various organs shares similar pathways, such as loss of the tumor suppressor p53, it makes sense from a biological perspective to design a genetic test that detects multiple cancers. On the other hand, arguments can be made to screen for cancers in single organs. Screening tools have only proven to be cost-effective when applied in targeted screening populations: no guidelines recommend screening for breast cancer in men, or in women under the age of 40, unless a positive family history significantly increases the individual’s risk for breast cancer. When applying screening to at-risk populations, a test that detects a variety of cancers may prove less cost-effective than separate tests detecting individual cancers of interest for that specific population. Screening of rare and/or highly aggressive cancers is generally not cost-effective, due to its inevitably high false-positive rates and the short window of opportunity, respectively. Hence, it would be more...

    Show More
    Competing Interests: None declared.
  • RE: Detection and localization of surgically resectable cancers with a multi-analyte blood test

    This article provides very interesting information about the early detection of cancer using a blood test. The authors indicate a very appealing outcome when they state in the title that this test is associated with “detection and localization of surgically resectable cancers”. As a medical oncologist I therefore eagerly read the paper.
    Unfortunately what I see is that the Authors did not use the test to detect resectable cancers, but rather that they tested the use of their assay in patients with resectable cancers. This not exactly what I inferred from reading the title.
    I would like to know if there are data available on the possibility of using this test to diagnose occult cancers when still resectable.

    Competing Interests: None declared.
  • RE: Detection and localization of surgically resectable cancers with a multi-analyte blood test
    • Margaret Pepe, Full Member, Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle
    • Other Contributors:
      • Prentice Ross, Full Member, Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle
      • Mark Thornquist, Full Member, Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle
      • Patrick Bossuyt, Professor of Clinical Epidemiology, Biostatistics & Bioinformatics, Academic Medical Center University of Amsterdam, Amsterdam

    This article generated a lot of attention in the media. There were statements about our having made great progress towards the goal of an effective pan-cancer screening test. However, few in the cancer biomarker research community could argue that this study was adequately designed to answer basic questions about the potential value of CancerSEEK for screening.
    The authors themselves acknowledge a key limitation in the study design, namely that the cases were already diagnosed with cancer at the time of blood draw. There is as yet no evidence that the biomarkers can detect cancer before it is evident clinically. Indeed, some of the components of CancerSEEK are already known to be poor indicators of preclinical disease and have consequently failed to be adopted into screening practice.
    Another study design issue concerns the choice of controls who were vaguely described as a “healthy control cohort” with median age 55 and no known history of several disease conditions. Were their samples collected and stored in similar manners and over similar time frames as those for the cases? Could differences in marker values between cases and controls be attributable to differences in sample collection procedures rather than cancer? Might differences in physiologic states such as anxiety or inflammation be a source of marker discrepancies between cancer and non-cancer subjects?
    There are questions also about bias in CancerSEEK performance statistics, as a result of us...

    Show More
    Competing Interests: None declared.
  • CancerSeek Test, is atractive but with serious shortcomings
    • Jairo Echeverry-Raad, Pediatrician, Head Professor, National University of Colombia, Faculty of Medicine

    Dear Editor,
    I really was enthusiastic with an elegant paper of Cohen et al (1) , published in First Realease Science paper in January 19/2018, about of new multi-analyte blood test (CancerSeek) for the detection and localization of surgically resectable cancers like a “fluid biopsia”. Having a rapid, easy to perform and relatively cost-effective test for the eight most prevalent types of cancer its advance attractive on oncological diagnosis.

    The technical procedure to obtain the CancerSeek test is very ingenious. For this aspect, as well as for all the splendid technology deployed in the fields of molecular biology that the study holds, I have no objections and my congratulations.

    However, with all respect, I have some critics from the clinical epidemiology and diagnostic tests accuracy grounds and from physician perpective that work in clinics scenarios as well.

    This comments regarding mainly with the design of the study and in the patients selection, in the Gold Standar and test (CancerSeek) work-up and in the applicability of test in clinic grounds.

    First of all, there is empirical evidence en diagnostic test, that the poorest design in the validation of accurary and operative performance is the Case Control study with noncosecutive patient serie’s, that infortunately Cohen et al assembled in your study. About this, Lijmer et al (2) , showed that studies with methodological shortcomings may overestimate the a...

    Show More
    Competing Interests: None declared.
  • RE: Liquid biopsy: exciting but needs more transparency
    • Kehong Zhang, Medical editor, Ivy Medical Consulting & Editing

    On 18 January 2018, Science published a diagnostic study by JD Cohen and colleagues (1). The authors developed a blood test (CancerSEEK) to detect cancers prior to metastasis. Using 1,005 patients with a variety of surgically resectable cancers and 821 healthy volunteers, authors achieved 70% sensitivity and >99% specificity. This work has been considered a major breakthrough toward early detection of cancers.
    The findings are exciting. But similar to many other studies, credibility of the results rests heavily on the “blindness” of the study. I trust that the authors made reasonable efforts in this regards, but believe the study could have been much more convincing if sourcing and coding of the samples were carried out by a third party with full authority and public trust.
    I suggest that Science coordinate an “experiment”: arranging a set of encrypted blood samples to be tested by the authors in a truly blinded manner, and publishing the results. I further advocate that Science taking a lead in setting up a policy to require, as a condition for publication, author promise to verify the findings as deemed appropriate. Such an action would enhance journal impact. More importantly, the policy would represent a major step towards more transparent and credible science.

    1. J. D. Cohen et al., Science 10.1126/science.aar3247 (2018)

    Competing Interests: None declared.
  • RE: Detection and localization of surgically resectable cancers with a multi-analyte blood test
    • Eleftherios Diamandis, Professor & Head, Division of Clinical Biochemistry, Mount Sinai Hospital, University Health Network, University of Toronto

    Cohen et al. presented compelling evidence that combination of protein based blood cancer biomarkers and circulating tumor DNA (ctDNA) can be used as an effective non-invasive test for early cancer detection. If replicated, these findings may have major implications in the battle against cancer.
    All mentioned protein cancer biomarkers were discovered 30 to 50 years ago and have already been thoroughly tested for early cancer diagnosis with hundreds of thousands of patients, either alone, or in combination with imaging. Current practice guidelines recommend against using these proteins, including combinations, for early cancer diagnosis due to sensitivity and specificity concerns. (Cancer Prev. Res. 2011; 4: 365-74).
    ctDNA is a new biomarker that has utility for various clinical applications, including early detection. A recent paper by another group provided evidence that ctDNA could be an early cancer diagnostic modality (Sci Transl Med 2017; 9:403). Our detailed analysis has shown that ctDNA sequencing technologies (liquid biopsies) are unlikely to detect early and asymptomatic cancer mainly due to insufficient amount of ctDNA in the circulation. In short, a 10-mL blood draw will yield less than one cancer genome equivalent, if the tumor is less than 1 cm in diameter or 1 g in weight (F1000R 2017;6:2129) making genetic analysis very challenging or impossible.
    We will not know if the data of Cohen et al. is translatable, until the 99% specificity is conf...

    Show More
    Competing Interests: Dr. Eleftherios P. Diamandis holds a consultant/advisory role with Abbott Diagnostics. Clare Fiala has nothing to declare.
  • Am I Lucky?
    • Mesut Tez, surgical oncologist, Ankara Numune Hospital Department of Surgery

    More than 2 years ago, Tomasetti and Vogelstein brought that question to prominence when they tried to sort out environmental versus inherited causes of cancer. They examined the extent to which stem cell divisions in healthy cells—and the random mutations, or “bad luck” that accumulate—drive cancer in different tissues(1).
    Vogelstein and Tomasetti developed new test that successfully detected cancers 70 per cent of the time in trials and they conclude that "... Our study lays the conceptual and practical foundation for a single, multi-analyte blood test for cancers of many types. We estimate the cost of the test to be less than $500, which is comparable or lower than other screening tests for single cancers, such as colonoscopy. The eight cancer types studied here account for 360,000 (60%) of the estimated cancer deaths in the U.S. in 2017 and their earlier detection could conceivably reduce deaths from these diseases. To actually establish the clinical utility of CancerSEEK and to demonstrate that it can save lives, prospective studies of all incident cancer types in a large population will be required. ...(2)" now!
    It is not easy to understand how cancer became an easily detectable disease from bad luck?
    Briefly, can we say; we are LUCKY if we have $500?

    1. C. Tomasetti, B. Vogelstein, Variation in cancer risk among tissues can be explained by the number of stem cell divisions. Science 347, 78-81 (2015).
    2. J. D....

    Show More
    Competing Interests: None declared.
  • RE: 0 REPORT Detection and localization of surgically resectable cancers with a multi-analyte blood test

    This article deals with one of the most long-standing, and largely unsuccessful to date, attempts of early detection and localisation of malignancy. In fact, the hopes to have identified a reliable maker of cancer via blood test were already high in the '70s.
    For instance, around that time a multitude of papers had been published about the C.E.A. has a "reliable index" for colonic cancer. Many other "markers" were identified in the 80's, but their clinical utility particularly in terms of screening and early diagnosis was later on disproved.
    The Authors do refer to some cancers, such as the ovarian cancer, were we do already have more reliable, albeit not reliable enough, markers such in the case of the ovarian cancer.
    In premenopausal patients, CA 125 sensitivity has been calculated at 64% whereas specificity 94.12%, positive predictive value 90% and negative predictive value 94.12%. Among postmenopausal women CA 125 sensitivity is thought to be 88.73%, specificity 98.07%, positive predictive value 98.44% and negative predictive value 86.44%.The list continues.
    It may be worthwhile to highlight that cancers may have endocrine and metabolic manifestations, and those may be overlooked, yet could be checked upon.
    Ultimately, it is questionable the clinical utility of a test that, to the question "do I have cancer?" allows a reply, if positive, "you may...perhaps fifty-fifty chance". A coin flip w...

    Show More
    Competing Interests: None declared.

Navigate This Article