COMMENTARY

'Twitter Cheerleaders': A Megaphone for Bad Science

Vinay Prasad, MD, MPH

Disclosures

September 19, 2018

A new class of physician has emerged in the era of social media. I call them "Twitter cheerleaders." A Twitter cheerleader is someone who bestows lavish praise on a new cancer drug, paper, trial, or test, while an objective and impartial look at the data may suggest caution or, at a minimum, less exuberant language. Cheerleaders uncritically parrot drug company marketing. How can readers recognize cheerleaders? Here are some of my tips.

1. Cheerleaders celebrate press releases that don't contain information on the magnitude of benefit. For example, "Wow, Pharma just announced that cost-a-fortune-amab met its primary endpoint of PFS in lung cancer! Great news for patients!" Upon closer inspection, the press release says a statistically significant P value was achieved but there's nothing about the size of benefit, toxicity, cost, or whether survival was improved. With all due respect, it might be prudent to wait for actual information before popping the champagne.

2. Cheerleaders celebrate a paper without reading it. It is always difficult to know whether someone read a paper before tweeting about it—unless, of course, you work for the NSA or Facebook—but there is a simple rule of thumb: If a paper first appears online on the New England Journal of Medicine website at 5:00 PM Eastern Standard Time, and someone tweets "Amazing work and fantastic results for patients" at 5:01 PM, it is probably fair to say that they didn't give it the ol' college read.

Cheerleading is not the same as optimism.

3. Cheerleaders call a drug that does not improve overall survival "miraculous." I know there is a lot of debate regarding the importance of surrogate endpoints (response rate and progression-free survival) and how often they should be used for drug approval, and I cannot settle that debate here. But when someone pulls out the big-gun words of cancer medicine—miracle, game-changer, revolution, or cure—and the published manuscript contains a Kaplan-Meier overall survival curve where the curve of the new drug and the control arm are superimposable, then I think it is fair to say that this word choice is inappropriate. Miracle drugs must improve survival, at the least.

4. Cheerleaders think any drug, no matter how marginal or ineffective, will be useful and it is just a matter of patient selection or some yet-to-be-discovered combination. If a drug improves survival 10 days in metastatic cancer[1] or has a < 10% single-agent response rate,[2] cheerleaders tend to say, it "may have an important role" or "may work better in combination" or that it is just a matter "of finding the right biomarker." All of these statements, like any speculation about the future, are possible, but useful biomarkers and effective combinations don't just fall from the sky. And often they don't materialize.

5.Cheerleaders re-tweet cost-effectiveness analyses sponsored by the maker of the cancer drug that show favorable results. "Wow, would you believe that? Even though cost-a-fortune-amab is expensive, it has been proven to be cost-effective in this elegant analysis by Pharma!"

6. Cheerleaders make frequent, lavish, hyperbolic claims about the future. Of course, many (myself included) believe that the future will be better than the past, but most of us are not fortunetellers. On any given Sunday, a cheerleader is more likely to say, "Future drug YXD-402 may be a game-changer!" than make mention of the reality that "an analysis of 71 drugs for solid tumors shows the average survival benefit was 2.1 months."[3]

7. Cheerleaders point to improvements in 5-year survival statistics as a measure of how good cancer drugs have become. If you think that stage-specific 5-year survival is a good judge of novel drugs, I would suggest Googling "lead time bias," "stage migration," and the "Will Rogers phenomenon."

Someone who meets merely one of these hallmarks is probably not a cheerleader; it takes consistently hitting several of them to earn the title. And cheerleading is not the same as optimism. Oncologists tend to be an optimistic group, and I consider myself an optimistic person. Cheerleading is irrational, overpromising, and unrealistic. Cheerleaders parrot drug company marketing. They often have a poor grasp of statistics and bias.

Two decades ago, James Watson made a famous blunder, claiming that Judah Folkman and angiogenesis inhibitors would "cure cancer in two years."[4] Andrew von Eschenbach, previous National Cancer Institute director, famously gaffed, claiming that we would "eliminate suffering and death" from cancer by 2015.[5] Neither of these prophecies materialized. But Twitter cheerleaders have escalated and magnified this problem. They turn the overstatement of the decade into the overstatement of the day.

Instead of cheerleaders, what we desperately need from physicians are realistic appraisals of what has been accomplished. Clear statements of future cancer goals. Cancer research funding policies that are driven by evidence, and greater honesty daily in the examination room and on the wards. Take off the rose-tinted sunglasses with the Pharma logo, and see the world with your own eyes. We would all be better off with fewer cheerleaders and more true scientists.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....