Cancer Therapy Advisor asked Keith T. Flaherty, MD, the editor-in-chief of Clinical Cancer Research, about the decision to publish the findings the way they were published, despite the issues with statistical significance. He said that the journal publishes research that spans a spectrum of both what he called “statistical tests” and “interpretive statements.”
This editorial approach appears to shift the burden of interpretation and assessment of results onto the reader.
“Statistical significance, borderline statistical significance, and trend are all terms that are used commonly in scientific publications,” Dr Flaherty wrote in an email to Cancer Therapy Advisor. “For analyses that are exploratory in nature and for which a prospective sample size was not calculated, with statistical power assessments, it is common to report numerically different observations in support of a hypothesis. Readers can choose to focus on statistical significance at the 95% confidence levels and disregard all other observations if they like.”
Dr Sainani also deemed it problematic that the overall survival data had been placed in a separate supplement, even though overall survival is superior to disease-free survival as an endpoint.
Cancer Therapy Advisor asked Dr Flaherty about the rationale behind not incorporating the data on overall survival in the main paper. He cited the inability to include all of the data in the body of most papers published in the journal. “In those cases, we are always confronted with making choices of priority regarding which tables and figures are more or less important,” he wrote. “Those decisions are made jointly by the authors and reviewers, with oversight provided by our editorial leadership.”
The most likely explanation of the findings about the alleged improvements that the researchers reported was that in older people, a target different from VEGF was likely the main driver for angiogenesis. “If they are undergoing angiogenesis because of some other factor, then obviously [bevacizumab] is not going to work,” Dr Weeraratna said.
The researchers knew that sFRP2 was increased during aging and that previous research had pointed to the protein as a driver of angiogenesis. Further research conducted as part of the new in vitro study and a study in mice found that the protein took over as an angiogenesis driver in older tumors. “And so, that’s why the tumors were not responding to Avastin,” she reasoned.
Dr Weeraratna thinks that the results might help to make treatment decisions based on patients’ age. “It tells us that we might be able to give Avastin to younger patients after all, instead of throwing the baby out with the bathwater,” she said. “It also tells us that finding ways to target angiogenesis beyond VEGF is very important, and that’s probably going to be the case for multiple cancers.”
However, Sunandana Chandra, MD, MS, assistant professor in the division of hematology oncology at Northwestern University Feinberg School of Medicine in Chicago, Illinois, who was not involved in the study, said that it is too early to apply these findings to clinical practice. “Although the study is intriguing in suggesting that VEGF inhibition with bevacizumab may be age dependent in melanoma patients, I don’t think the data [are] robust enough to impact current clinical practice,” she wrote in an email. “Studying this formally in the setting of a clinical trial would be the next step.”
Wasif M. Saif, MD, deputy physician-in-chief and medical director at Northwell Health Cancer Institute in Lake Success, New York, who was not involved in the study, wrote in an email that more research is also needed to examine the validity of sFRP2 as a potential diagnostic or prognostic biomarker in certain types of cancer.
Dr Sainani concluded that while the findings suggest that patient age in this disease setting could be worth further examination in prospective clinical trials, “the evidence provided by this study is quite weak,” she said.
Disclosure: Some of the authors of the discussed study disclosed financial and non-financial ties to the pharmaceutical industry. This includes one author who disclosed a non-financial link to Roche Pharmaceuticals. For a full list of disclosures, please refer to the original article.
The original paper also included the following statement in the “Acknowledgements” section at the end: “The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.”
- Corrie PG, Marshall A, Nathan PD, et al. Adjuvant bevacizumab for melanoma patients at high risk of recurrence: survival analysis of the AVAST-M trial [published correction appears in Ann Oncol. 2019;30(12):2013-2014] [published correction appears in Ann Oncol. 2019;30(12):2013-2014]. Ann Oncol. 2018;29(8):1843-1852. doi:10.1093/annonc/mdy229
- Corrie P, Marshall A, Lorigan P, et al. Adjuvant bevacizumab as treatment for melanoma patients at high risk of recurrence: Final results for the AVAST-M trial. J Clin Oncol. 2017;35(15_suppl):9501-9501. doi:10.1200/JCO.2017.35.15_suppl.9501
- Fane ME, Ecker BL, Kaur A, et al. sFRP2 supersedes VEGF as an age-related driver of angiogenesis in melanoma, affecting response to anti-VEGF therapy in older patients. Clin Cancer Res. 2020;26(21):5709-5719; doi:10.1158/1078-0432.CCR-20-0446
- Sainani K. Misleading comparisons: the fallacy of comparing statistical significance. PM R. 2010;2(6):559-562. doi:10.1016/j.pmrj.2010.04.016
- American Association for Cancer Research. Press Release. Response to adjuvant bevacizumab among patients with resected melanoma may vary by age. Published October 23, 2020. Accessed December 5, 2020.