You might have heard that many new cancer medicines offer little benefit. This month, the BMJ published a review finding that for most cancer drugs approved by the European Medicines Agency between 2009 and 2013, there was neither published evidence that they extend overall survival nor improve patients’ quality of life. In 2015, JAMA reported similar observations for oncology drugs approved by the U.S. FDA between 2008 and 2012. Both papers focused on evidence from randomized clinical trials.
These agents are all costly, besides. Why treat cancer? Some still ask.
After 1990, when I was a first-year fellow in oncology, the U.S. death rate from cancer fell by over 25%. As I outlined in this year’s ASCO educational book, most advances are incremental and imperfect. While we’re nowhere near where we should be, or could be—in either preventing or treating cancer—science has led to evident progress: Mortality from cancer fell steadily between 1991 and 2014. This favorable trend occurred despite a synchronous rise in the overall incidence of invasive cases and suggests that many treatments are indeed helpful.
One possible explanation is that all measured gains against cancer, after surgery and radiation, come from the powerful activity of a few blockbuster drugs like Rituxan in lymphoma and Gleevec in chronic myelogenous leukemia and Herceptin in HER2 positive breast cancer. A more likely scenario is that some medications help many patients a lot, and so it’s easy to document their value, while most agents help fewer patients to a varying degree, which is messier and harder to prove in clinical trials, but no less true.
Perhaps the literature fails to capture the clinical benefit of many oncology drugs. This could happen for several reasons. One is the under-appreciation that progression-free survival (PFS) can be truly helpful, permitting many people with advanced cancer to more comfortably go about their lives. Another is that randomized clinical trials (RCTs) may not be designed or statistically powered to reveal true benefits in small fractions of patients.
Let’s consider progression free survival. PFS refers to the time before a cancer progresses. For many cancer drugs, PFS is assessed in clinical trials and reported years before the long-term impact on overall survival (OS) can be known. A very recent example is durvalumab (Imfinzi). In a randomized trial of lung cancer patients with stage 3 tumors that could not be removed surgically, after radiation and standard chemo, durvalumab prolonged PFS by 11 months compared to a placebo. Based on this trial result, some say it’s the new standard of care.
As a medical student, I was taught to weigh overall survival more than PFS and other “surrogate” measures of a treatment’s value. In general, I still do. But I’ve since learned that for many people living with advanced cancer, no progression is desirable and good. Or, at least, it beats the alternative: Many patients would choose stable disease over expanding tumors.
Sometimes, a cancer drug’s full benefit doesn’t emerge for years. After it’s evident that an agent has activity against a tumor type and is generally safe, optimizing how it’s given, managing side effects, and dosing in more effective combinations can take a long time. For patients with advanced cancer that is growing unstoppably, who are well enough to receive treatment and wish to do so, waiting for results of RCTs to confirm a promising drug's overall survival benefit could be a fatal mistake.
You might remember the spring of 1987, when AZT was approved in a hurry. The first anti-HIV agent turned out to be an imperfect drug; I don’t think it cured anyone. But it helped some HIV-infected patients to live a bit longer. Some of those early AZT-taking patients survived long enough to try another drug. Some of those patients are still around, today. At the time, we doctors wouldn’t have thought this possible. There is a lesson in this story for cancer patients, I believe.
Recently I attended a talk on HER2+ breast cancer by Dr. Eric Winer and was reminded of why it’s worth paying attention to drugs that extend PFS in aggressive malignancies. He spoke about how trastuzumab (Herceptin) flipped the outlook for patients with HER2 positive (+) tumors. In the 1990s, from his perspective as a medical oncologist specializing in breast cancer, “it seemed like HER2+ breast cancer was just about the worst form you could have.” But this is no longer the case, he said. Herceptin changed everything. HER2+ disease—what was a very aggressive and hard-to-treat form of breast cancer, that tended to relapse early and spread to the brain—can in many cases now be managed for years, like a chronic disease.
He noted that in the first major report on trastuzumab, a 2001 publication in The New England Journal of Medicine, only a very modest benefit was observed: a 3-month extension of PFS, along with significant toxicity, in a randomized trial of women with metastatic HER2+ breast cancer. Back then, the advantageous effects of this antibody were not clear. Pathologists didn’t yet know how to accurately measure HER2, and so preliminary trials included tumors that were unlikely to respond. Initially there was a high rate of cardiac toxicity. Allergic reactions occurred. Over time, doctors learned how to minimize those problems.
As I’ve considered previously, old-school randomized controlled studies of cancer drugs take years to complete. They cost lots—tens or hundreds of millions of dollars, each—including salaries of cancer investigators and staff including administrative support, and by the time they’re done and analyzed there is often a new and possibly better or less toxic drug available, which most patients would want to consider. Pharmaceutical companies pay for many of the big studies, which results in there being essentially no modern head-to-head testing of similar agents or drugs that might be used to treat one condition. Apart from industry, RCTs may involve wrong doses, archaic or otherwise-flawed comparators, evaluated in selected patient populations who tend to be higher-educated and disproportionately receiving care at university hospitals and cancer centers. Despite so much investment, RCTs don’t always yield “real-world” information that’s applicable to many patients who want the best drug available for their precise tumor.
From the perspective of an individual patient, a medication may be beneficial even if it doesn’t extend life. Consider the hypothetical predicament of a 55-year old man with advanced liver cancer that presses on his bile duct and causes him to experience pain, jaundice and vomiting. That same liver cancer might irritate his diaphragm, leading to constant hiccups or coughing fits. If his tumor could be controlled enough, so that it doesn’t crush the bile duct and involve the lower chest, he might feel better while he’s still alive, which could be for a year or two. There are many similar examples: cancer in the lungs causing windedness and a bloody cough; cancer in the bones causing excruciating pain; cancer in the abdomen causing uncomfortable bloating and nausea. You get the idea.
In each instance, a cancer-slowing agent might help a person to feel well, go about his or her activities, and enjoy more days. There’s always a tradeoff; toxicity needs be weighed, among other considerations. Meanwhile a better drug in the pipeline might be approved or become available in a clinical trial, and the patient might try that and possibly continue to live, longer.
Some call this the hitchhiker model of cancer survivorship. However intriguing this concept seemed in 2008, when I first read about it, this idea has become more relevant each day, as new drugs emerge and scientists figure out better ways to evaluate and treat many kinds of cancer.
This is not a theoretical consideration. It’s happening now. The problem, from my point of view as a patient advocate, is that it’s not happening fast enough. Potentially helpful drugs are being kept out-of-reach, by price or by lack-of-approval, from many people who are alive with advanced cancer and who wish to try these new agents.
Given the extensive news coverage of the BMJ report, it’s worth mentioning that it wasn’t until recent years that oncologists began routinely assessing quality-of-life or patient-reported outcomes in randomized trials. The authors of that study and an accompanying editorial highlighted the absence of evidence that most drugs approved in Europe through 2013 improve quality-of-life. But that information was absent does not mean that new cancer agents don’t improve patients’ well-being or cause net harm. It means only that there is a need for more information about how patients feel.
As for how to pay for these drugs, this is a serious and growing concern. But it’s a separate issue from whether or not treatments are helpful and worth trying. The solution will involve better matching of medications to likely positive responders, such as by companion diagnostics to predict who will probably benefit. But for many cancer drugs in 2017, those markers are not yet optimized. In my view, many off-the-shelf drugs needn’t cost so much, but that’s a topic for another day. For the time being, I do think much of the responsibility falls with doctors, who need to use judgment in advising patients and prescribing new cancer drugs wisely.
Years ago, I watched colleagues who treated bedbound patients, some older and many near death, with toxic chemotherapy for which the patient didn’t quite consent. Some oncologists didn’t know when to stop, or lacked the skill or motivation to tell patients when anti-cancer therapy was no longer appropriate. Those treatment circumstances should be history. Whoever are the patients who do try new cancer drugs, on trials or otherwise, they should be aware of the risks and benefits, and well enough to take them, volitionally.
