Patient access to new cancer treatments — an uncertain and fragile process
“What I quickly learned after my diagnosis is that the world of a cancer patient has many parts and a good deal of uncertainty.”
— Tom Brokaw
As Tom Brokaw eloquently stated, uncertainty is at the heart of the cancer experience for our patients.
It impacts many aspects of their lives, from their ultimate survival and treatment options to their financial, social and emotional well-being, as well as that of their families and caregivers, and much more.
For many patients, uncertainty surrounding access to the best possible treatment can drive decisions for second opinions, referral to highly specialized treatment centers and participation in clinical trials.
Uncertainty in different domains also is a central feature of our role as oncologists — the use of clinical trials, especially in the phase 3 setting, is based on the goal of addressing uncertainty and coming up with definitive answers on the effectiveness of new therapies. Informed by the data from clinical trials of various designs, responsibility for granting our patients access to these new treatments belongs to regulatory bodies and, subsequently, payers.
That said, ultimate responsibility for the use of these new treatments lies with us as prescribers, and we should be held accountable for their effectiveness and the costs incurred by our practice.
Still, making good judgements about true effectiveness is very difficult. We function in a specialty in which meaningful benefits in terms of survival, symptom control or quality of life can be difficult to discern from the results of clinical trials — either because the improvements in outcome are marginal, or the right questions simply were not asked in the context of the trial.
Phase 3 fragility
Early FDA approval of promising new therapies has been an important and positive step to remove potential obstacles in patients’ access to new treatments but also has brought new challenges for us as oncologists and for the health care system.
One frequent observation has been that new agents have been approved based on statistically significant differences in trial endpoints, with P values reaching the < .05 standard, when the outcome differences are clinically meaningless, with only minor (often measurable in a few weeks) improvements in time to events.
As I learned many months ago and described in a previous editorial, the P value is not the gold standard many of us thought it was — it is subject to interpretation and to nuances in statistical method that I can’t pretend to understand. Evidence for this comes from studies that have assessed the clinical benefits of cancer drugs approved by the FDA against frameworks such as the European Society for Medical Oncology Meaningful Clinical Benefit Scale or the ASCO Value Framework. Many agents that have reached the threshold for approval have been shown to have minimal clinical benefit.
A study published last month in The Lancet Oncology extends this observation and reports on the concept of “fragility” of phase 3 trials.
I’ll confess, this was a new concept to me — another example of lifelong learning — but it is quite elegant. Fragility measures how easy it is to flip the results of a randomized clinical trial from significant (P < .05) to nonsignificant by moving events from one arm to the other. In other words, in a positive clinical trial with superior outcome (fewer events) in the experimental arm, fragility measures how many events must be moved from the control to the experimental arm to render the difference nonsignificant.
The authors took available results from 17 of 36 phase 3 trials which led to FDA approvals from 2014 through 2018 and showed that about half these studies had very low fragility indices, meaning the P values were driven by very small numbers of events. Of concern, in some cases, the number of events required to define the studies as fragile was less than the number of patients lost to follow-up.
In short, this study provides further evidence that we should be circumspect about the real benefits of new treatments and should assess these against established value frameworks, which ideally also factor in treatment costs.
Uncertain true value
On the subject of cost, an email arrived in my inbox this morning containing a market analysis of what it described as the “gold rush” of T-cell immunotherapeutics for cancer.
It anticipates an annualized growth rate of over 40% for the next decade and reports on the billions of dollars already invested in this emerging and exciting area. This market-based analysis of a potentially transformative new treatment clearly illustrates the high-stakes nature of modern oncology for those fueling the engine of research and development, but it doesn’t address the group for whom the stakes are highest — patients.
Early experience with T cell based strategies has been remarkable, but so have the costs. Because our patients, either directly or indirectly, will bear part of this cost, we need to understand what we are getting ourselves and our patients into when recommending these approaches.
A study published in June in Journal of Clinical Oncology tried to evaluate the cost-effectiveness of chimeric antigen receptor T-cell therapy for adults with relapsed aggressive B-cell lymphomas and compared projected costs with those of the current alternative strategy — namely salvage immunochemotherapy followed by stem cell transplantation.
The authors concluded that based on 2018 pricing, CAR T-cell therapies could meet the threshold of less than $150,000 per quality-adjusted life-year. An editorial that accompanies this publication pointed out that the estimates of benefit are based on very optimistic assumptions regarding the long-term outcomes of CAR T-cell therapy in this disease. A relatively small negative variance in the actual outcome over time would substantially reduce the cost benefit for this treatment.
Although the study represents a rigorous attempt to quantify the potential benefit of CAR T cells, it fails at one level because the effectiveness data simply aren’t there, and probably won’t be for a few more years. The uncertainty regarding the cost-effectiveness of this treatment will remain until randomized data are available from ongoing studies and, even then, the P values may not tell the whole story.
For the moment, the only strategy certain to improve the cost-benefit ratio would be to reduce the cost — interestingly, in other countries, notably the U.K., regulatory bodies were able to negotiate an undisclosed reduced price for CAR T products, which has allowed them to approve these products for relapsed DLBCL.
Of course, this still involves assumptions about long-term benefit that may or may not be true, but it would be interesting to know how the European and U.K. pricing of CAR T cells would have affected this cost-effectiveness analysis. Even with a lower cost, one has to conclude that this analysis is premature given the lack of long-term data.
Emerging new cancer treatments inevitably come with uncertainty about their true value. We need to remain vigilant to the sometimes-shaky foundations on which approvals are based, either because the marginal benefits do not stand up to robust evaluation, or because analyses are conducted in the absence of mature data.
We have a natural tendency to default to promising new agents for our patients, but we need to be aware that study results can be as uncertain and fragile as our patients sometimes are.
For more information:
John Sweetenham, MD, FRCP, FACP, is HemOnc Today’s Chief Medical Editor for Hematology. He also is associate director for clinical affairs at Harold C. Simmons Comprehensive Cancer Center at UT Southwestern Medical Center. He can be reached at firstname.lastname@example.org.
Disclosure: Sweetenham reports no relevant financial disclosures.