News Stories and Medical Breakthroughs

A recent AACC SmartBrief1 included a link to a news story from the Medical Device and Diagnostic Industry newsletter reporting a new breakthrough for breast cancer diagnosis, with the catchy title: Blood Test Can Detect Breast Cancer 5 Years Before Symptoms Occur (1). The new test, a pilot study, was presented at the annual conference of the National Cancer Research Institute (NCRI) of the United Kingdom during the first week of November 2019 and has not yet been formally published after peer review. Nevertheless, it represents a public disclosure. The group from the University of Nottingham has published original work in this area in the past, as discussed below. Their new test is based on analysis of a panel of serum autoantibodies against tumor-associated antigens. JALM readers are quite familiar with news stories of this type, which promise to revolutionize cancer diagnosis and treatment with new technologies. If even a fraction of these discoveries proved to be successful, cancer and other disease burden would be minimal! However, as it usually happens, the stories persist for a few days and then they go into oblivion. This problem is particularly true for cancer diagnostics (2). How would a lay person know if the announced breakthroughs are real or relevant? This question is not easy to tackle, as it is sometimes difficult to determine discovery significance at the time of publication/public disclosure/press release or whether the results will stand the test of time. Recently, referring to the related issue of scientific irreproducibility, we proposed a 5-year reflection for high impact publications and announcements. A high-impact publication is arbitrarily defined as a publication appearing in a high-impact journal (e.g., impact factor >15) or receiving a high number of citations (e.g., >50 per year). Briefly, we suggested that seemingly major discoveries should be re-examined at 5 or 10 years to see what happened to them (3, 4,). This can be done with a published commentary from the original authors, as we describe in detail elsewhere (3, 4). Another effective strategy would be for experts in the field to issue counter-press releases, critiquing the announced data. However, this approach is unrealistic since not many scientists have the time, nor the desire, to enter into discussions of this kind. The public will likely get more confused than enlightened from such scientific debates. However, in an effort to contribute in this area, we will analyze the data described in the news story, to illustrate the point that many apparent new breakthroughs merely rehash previous knowledge, sometimes many years old, with incremental improvements. The technology described in the news story dates back to 1995-96 when Robertson et al., of the University of Nottingham, published an article describing the value of serum p53 autoantibodies for breast cancer diagnosis (5, 6,). In fact, our group had already published the same concept a few years earlier, in 1992 (7). Our 1994 publication was one of the most extensive investigations of the use of serum autoantibodies as cancer diagnostic markers, and we reported the sensitivities and specificities of these autoantibodies in various cancers by analyzing 1393 specimens from cancer patients and 230 specimens from controls (8). Since then, this work has been cited over 265 times. We found the highest autoantibody sensitivity for ovarian and colon cancer (15%) while for breast cancer, the sensitivity was 5%. In that article, we mentioned that the discovery of additional autoantibodies may improve the low diagnostic sensitivity, but this advancement would likely come at the expense of specificity. This work was subsequently expanded by other groups using newer techniques such as antigen microarrays for detecting breast and other cancers. This included investigators funded by the Early Detection Research Network (EDRN), an NIH/NCI Organization with the mission to discover biomarkers for early cancer detection (9,). The best combinatorial markers for breast cancer achieved sensitivities in the 30% range, with a specificity of 98%. A recent review of serum autoantibodies for cancer diagnosis has been published (10). In the newly announced work, the reported sensitivities ranged from 29% to 37%, with specificities in the range of 84%–79%, respectively. Are these values sufficient for a clinical test used to identify asymptomatic patients with breast cancer? This application is better known as screening. Below, we will provide an example, using typical values of 35% sensitivity and 80% specificity, to illustrate our points. For simplicity, other possible complications of breast cancer screening, such as the detection of indolent cancers (which grow slowly and do not need treatment), will not be considered. Let us assume a population of 10 000 asymptomatic women is screened, of whom 100 have breast cancer (1% prevalence) and 9900 women do not have breast cancer (designated as normal). Before any test is administered, the chance of these women having breast cancer is 1% (equal to the prevalence) and the chance of them not having cancer is 99%. After the test is performed, 35 women with cancer (100 x 0.35) will be positive for the test (true positives, TP), but 1980 normal women (9900 x 0.20) will also receive a positive result (false positives, FP). On the other hand, 65 women with cancer (100-35) will receive a negative result (false negatives, FN) and 7920 normal women (9900-1980) will be negative (true negatives, TN). Based on these numbers, we can calculate the positive and negative predictive value (PPV and NPV) of this test: PPV = 35/(35 + 1980) x 100 = 1.7%; NPV = 7920/(7920 + 65)x100 = 99.2%. These data show that such a test for breast cancer screening would be useless. If the test is positive, it increases the chances of a woman having cancer from 1% (before the test) to 1.7% (after the test), a negligible increase. The same applies to the women who were negative; their chances of having cancer decrease negligibly from 1% to 0.8%. We conclude that the featured test in the news story is neither new, nor necessarily effective for early detection of asymptomatic breast cancer. Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. Authors’ Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest: Employment or Leadership: None declared. Consultant or Advisory Role: E.P. Diamandis, Abbott Diagnostics. Stock Ownership: None declared. Honoraria: None declared. Research Funding: None declared. Expert Testimony: None declared. Patents: None declared.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.