Not sure how much more breathless enthusiasm I can take
Each month, John Sweetenham, MD, and I review a stack of reports from conferences and the international literature for reporting in HemOnc Today.
Perhaps I’m just becoming old and crusty, but I’m really growing tired of the increasingly inaccurate or irrationally breathless enthusiasm of many new-age presenters of clinical trials when describing their work.
What is so disappointing is the progression of steadily less critical analyses of clinical trial data.
I understand that OS comparisons can be confounded in diseases where multiple lines of treatment are available, and also that there are now many cancers that can benefit from multiple lines of salvage treatment.
However, that doesn’t mean OS should be ignored as a seminal endpoint.
The usual excuse is that it is too early in the trial to evaluate OS and the presenters have rushed to present their new study armed only with DFS or PFS data ... but that is exactly the point. What’s the rush?
If one reviews the series of studies reported these days in the issues of HemOnc Today and all similar broadsheets, they are littered with interesting but dreadfully preliminary reports with exactly that problem.
Compounding the felony is the proliferation of less robust endpoints. We now routinely see DFS or PFS as a standard endpoint, with excited reports of significant P values (or HRs, forest plots, waterfall plots or swimmer plots) that reflect increments of median survival measured in days or weeks.
Perhaps my pet hate is the knucklehead trialist who reports “stable disease” as a response category without bothering to identify measurable progressive cancer prior to initiating treatment with the latest therapy. The claim of stable disease as a response outcome for a tumor that is not progressing is a pretty easy (but useless) win.
When one discusses treatment honestly with a patient with advanced cancer, it is hard to raise enthusiasm for a survival increment from a new treatment of only 2 months — “Yes, Ms. Smith, I know you are only 30 years old, but this new (and expensive) therapy has a great P value and could increase your survival by 2 months!” That phrasing really doesn’t have a ring to it, yet it is implied in so many of the novel targeted and immuno-oncology therapeutic reports in recent times.
‘Down migration’ in control arms
One of the truly odd features of recent oncology is the down migration of DFS or PFS in the control arms of randomized trials.
One would think that with all our progress in diagnosis, stage migration and supportive oncology, standard treatment arms would be looking better, and thus DFS, etc, would have improved, making it harder to show benefit from novel therapies in randomized or comparative studies.
Increasingly often, one reads of a 50% change in HRs showing a hike from median DFS of 3 to 4 months with standard chemotherapy up to the dizzy heights of 6 months, and that is seen somehow as a win.
I recall so clearly for many of the current subjects of randomized trials, such as breast cancer, lung cancer, untreated colorectal cancer, etc, DFS and PFS figures from the 1980s that far exceeded those being reported today.
So, what’s up? Is cancer becoming more aggressive, given that in the past 30 years stage migration (namely, increased testing and sensitivity of testing that shows ever-smaller deposits of cancer) should be associated with improved outcomes? Have we returned to the bad old days where conflict of interest influenced the interpretation of data and the reading of tumor measurements? Is it possible that trial clinicians are working less hard to leverage the therapeutic impact of standard therapies on control arms and thus producing worse outcomes inadvertently?
Next time you hear another breathless and hyped-up report of a new wonder drug at a meeting, or study the peer-reviewed results in a reputable journal, pull out a copy of Journal of Clinical Oncology or Cancer from the 1980s and check out the survival curves and PFS/DFS figures reported from equivalent studies of that time. Compare some of the reports emanating from ESMO Virtual Congress 2020 with study outcomes from 30 years ago.
‘Look more carefully’
I want to make one important point to set context. There is no question that the early seminal studies of targeted and immuno-oncology agents have had a huge impact on survival of renal cell carcinoma and malignant melanoma, and these advances are real and paradigm-shifting.
The point of this editorial is simply to remind us that well-applied conventional systemic therapy remains an effective option for many other solid tumors, and we must be critical in assessing reported outcomes in a range of solid tumor trials that are currently suggesting that targeted therapies and immunotherapies are huge breakthroughs in everything.
It is time to look more carefully at reports of DFS and PFS, to consider the true biological relevance of a treatment that “induces” stable disease, and to remember that offering an OS increment of less than 4 to 6 months is usually nothing to celebrate with excitement.
As this is my last editorial before the upcoming U.S. federal and state elections, and it is focused on truth and accuracy in reporting, I would be negligent if I did not applaud the editors of The New England Journal of Medicine for taking a stance supporting science and high-quality, discerning medical care.
For the first time in more than 200 years, the NEJM’s editors have departed from their nonpartisan stance, citing the importance of government’s role in leading a socially responsible attack against the menaces of COVID-19 and noting that the present federal government has turned a crisis into a national tragedy. This is a view that I completely support (see my Aug. 25 editorial, “The accidental ‘stupidpower’ — It is what it is”).
Kudos to these courageous medical editors and to all others who have taken a stance in trying to remind the population of the United States that there is no magic imbued in the current presidency and it is time for a change.
Editors. N Engl J Med. 2020;doi:10.1056/NEJMe2029812.
Join the conversation:
Share your thoughts on this editorial at healio.com/authors/draghavan.