It’s good to share, both in research and in life
“Share your knowledge. It is a way to achieve immortality.”
— Dalai Lama XIV
As I reflect on the ASCO Annual Meeting — specifically the new data presented and what I learned this year — I think of it as the ultimate intense “sharing” experience from an oncology perspective.
It is nearly impossible to tell whether the sharing of knowledge, experiences and insights at ASCO and other similar events has a measurable impact on the care of our patients, but I have to believe at some level we bring back information from the meeting that benefits their lives, as well as the lives of their caregivers.
The relationship between “sharing” in a very broad sense and impact for patients with cancer has been in the spotlight recently. My attention has been drawn to two very different aspects of this discussion — how we share our data and how we share our lives.
The concept of data sharing in cancer has gained new attention in the last few months, largely because of the national cancer moonshot initiative led by Vice President Joe Biden.
The issues of siloed information in electronic medical records (EMRs), poor interconnectivity of different EMR platforms, privacy issues related to protected health information, and the potential for “cancer politics” to inhibit sharing of information have all been highlighted, as has an undertaking to recruit a diverse group of experts and introduce legislative changes that will facilitate sharing of information about individuals with cancer.
The hope is that this will accelerate progress in cancer prevention, care and research. The leaders of the moonshot are banking on sharing of information and the development of “big data” sets as an effective part of the infrastructure to improve cancer outcomes.
Many individuals and organizations already share this vision and have developed their own data sharing platforms. Examples include the Oncology Research Information Exchange Network (ORIEN) and ASCO’s CancerLinQ project. These collaborations already have had early successes, including the recognition of previously unobserved drug toxicities and the rapid identification of eligible patients for early-phase clinical trials.
There also are emerging examples of other data-sharing strategies that likely will have long-term benefits.
For example, a study from Dana-Farber Cancer Institute and Blue Cross Blue Shield of Massachusetts combined clinical, administrative and claims data around end-of-life care. Researchers identified major differences in service utilization in patients with different types of cancer and a high rate of use of cancer services outside Dana-Farber — information that would otherwise not have been accessible to the cancer center. This kind of data sharing provides new insights into gaps in care coordination and an opportunity for quality improvement that otherwise would not be realized.
The issue of sharing data from clinical trials has provoked much discussion and editorial comment over the last few months.
The concept of sharing de-identified, individual patient data from clinical trials isn’t new; many meta-analyses have been conducted using such data and they have made important contributions to our understanding of cancer treatment. That said, the sharing of such data generally has not been systematic and has not been mandated, so opportunities for re-analysis or combined analyses of trial data have been limited. That is now changing.
Several industry sponsors of clinical trials have been submitting data to publicly available, open-access platforms for many years. The content and timing of the release of these data has been driven by company policy but, in general, the data have been made available once the study results have been published. Each of the open-access platforms has established a review process for applications to access these data and parameters to be met for publication of the “new” results.
A more stringent approach
Following an exhaustive report from the Institute of Medicine, a more stringent approach to data sharing from trials is now being proposed.
The International Committee of Medical Journal Editors recently announced that, as a condition of publication in one of its journals, it will require authors to share de-identified patient data from the article within 6 months of publication. It also proposed that the data-sharing plan should be a component of clinical trial registration — a practice already in place at clinicaltrials.gov — and that the data-sharing plan should be submitted with the manuscript.
Commentary on these proposals has highlighted potential benefits and pitfalls of this quest for transparency in trial conduct.
An excellent editorial in The New England Journal of Medicine cited a study in colon cancer in which new data on molecular markers of disease progression emerged as a result of data sharing.
The opportunity to gain new insights using previously published data in this way is a compelling argument in favor of this practice and — as has been pointed out by several authors — fulfils a responsibility that we, as clinical investigators, have to our patients. We fail in our responsibility to our patients when we open poorly accruing, poorly designed or underpowered studies.
Data sharing provides an opportunity for information from these studies to be useful to future researchers, extending the benefits far beyond those envisioned in the original trial design.
Not surprisingly, there are many concerns about this initiative, some of which relate to patient privacy, quality of data, as well as documentation and coding of data elements. In addition, the investigators who conduct subsequent studies may not understand many of the nuances of the original study design. There also are concerns about so-called research parasites who may use the hard-won data of previous investigators without adequate recognition of their original work.
I doubt if any of these issues are deal-breakers. I expect a transparent and generally acceptable process to be introduced to allow greater access to these data. The impact on patient care remains unknown.
Current data suggests the use of the available open-access platforms is quite low, and that only a small proportion of re-analyses have produced conclusions that differed substantially from the primary publication. It isn’t clear how many of these would have impacted patient outcome.
Another ‘sharing’ phenomenon
Changing gear, but still on the subject of patient impact, an intriguing study by Gomez and colleagues, recently published in Cancer, adds more evidence that who we choose to share our lives with may affect our cancer outcome.
There is a substantial body of literature that shows married patients experience better survival outcomes for most common cancers than unmarried patients; that the effect of marriage is greater in males; and that this effect tends to diminish with age.
Several hypotheses have been proposed to explain this difference. Many relate to associations with socioeconomic status, health behaviors, exercise and diet, and some have linked social support systems with biologic processes that may be protective against cancer through mechanisms such as neuroendocrine pathways.
Whatever the explanation, the phenomenon appears real, and this study concluded the differences are probably not related to economic disparities or to differences in insurance coverage. Something else seems to be at play.
So, we may not be able to define exactly what it is about a stable relationship that can affect cancer outcome, but this form of sharing seems to have a measurable benefit.
I’m quite sure the Dalai Lama was not thinking about cancer survival, and I realize I am quoting him completely out of context. However, it seems that even if sharing doesn’t give you immortality, it might improve survival.
Bauchner H, et al. JAMA. 2016;doi:10.1001/jama.2016.2420.
Ebrahim S, et al. JAMA. 2014;doi:10.1001/jama.2014.9646.
Gomez EL, et al. Cancer. 2016;doi:10.1022/cncr.29885.
Institute of Medicine. Sharing clinical trial data: Maximizing benefits, minimizing risk. 2015. Available at: www.nap.edu/catalog/18998/sharing-clinical-trial-data-maximizing-benefits-minimizing-risk. Accessed on May 15, 2016.
Longo DL and Drazen JM. N Engl J Med. 2016;doi:10.1056/NEJMe1516564.
Navar AM, et al. JAMA. 2016;doi:10.1001/jama.2016.2374.
Stuver SO, et al. J Pain Symptom Manage. 2016;doi:10.1016/j.jpainsymman.2016.02.004.
Taichman DB, et al. Lancet. 2016;387:e9-11.
For more information:
John Sweetenham, MD, is HemOnc Today’s Chief Medical Editor for Hematology. He also is senior director of clinical affairs and executive medical director at Huntsman Cancer Institute at University of Utah. He can be reached at firstname.lastname@example.org.
Disclosure: Sweetenham reports no relevant financial disclosures.