24 Scientific American, November 2019THE SCIENCE
OF HEALTHClaudia Wallis is an award-winning science journalist whose
work has appeared in the New York Times, Time, Fortune and the
New Republic. She was science editor at Time and managing editor
of Scientific American Mind.Illustration by Celia KrampienA Dilemma
with New Drugs
Most of the time we don’t know
if they are better than the old ones
By Claudia Wallis“New and improved.” These words have been yoked together in
so many marketing campaigns that we tend to accept them as
inexorably linked. But when it comes to new medications, don’t
swallow them without a healthy dose of skepticism. Many or
most new drugs are not—or at least not provably—an improve-
ment over the best existing drug for a given condition, and the
fast-track drug-approval processes that have prevailed in recent
years have added to the uncertainty about their advantages.
A recent report in the British Medical Journal, entitled “New
Drugs: Where Did We Go Wrong and What Can We Do Better?,”
offers an analysis of the issue. The authors looked at 216 drugs ap-
proved by German regulators between 2011 and 2017; 152 were
newly developed, and 64 were existing medications approved for
new uses. Only 25 percent of the medications were deemed as of-
fering a “considerable” or “major” advantage over the established
treatment (termed the “standard of care”), and 16 percent had a
minor or nonquantifiable advantage. Fully 58 percent had no prov-
en added benefit in terms of lowering mortality, reducing symp-
toms or side effects, or improving health-related quality of life.
“This doesn’t mean we are sure there’s no added benefit,” leadauthor Beate Wieseler said in an interview. “It just means we have
no positive proof. Either we have no studies at all [comparing the
new medicine with the standard of care] or we have studies, but
they aren’t good enough.” The record was “particularly egregious,”
she and her colleagues wrote, for drugs that treat psychiatric and
neurological disorders and those for diabetes, with only 6 and
17 percent, respectively, offering a confirmed added benefit.
Wieseler and her co-authors work for Germany’s Institute for
Quality and Efficiency in Health Care, which evaluates new treat-
ments and advises on whether the country’s health care system
should pay a premium for them. Such organizations, known as
health technology assessment (HTA) agencies, have become
“enormously more powerful” in many countries’ efforts to manage
the spiraling cost of new drugs, says Sean Tunis of the not-for-
profit Center for Medical Technology Policy in Baltimore. HTA
works a little differently in the U.S., he explains: “If payers think a
new drug is not any better than a drug that we already have, they
will do things like requiring you try the cheaper drug first.” Insur-
ers and Medicaid will often insist on this kind of “step therapy.”
Germany’s HTA is probably the most persnickety about de-
manding head-to-head trials to prove that a new treatment beats
the existing standard. This is not always practical. For one thing,
such studies can be hugely expensive and time-consuming, with
no guarantee of success. “What the authors are focused on is get-
ting new, differentiated medicines at a low cost, and what they
are missing is a sense of the complex economic underpinning of
developing new medicines,” says Ken Moch, president and CEO
of Cognition Therapeutics, a biotech firm in Pittsburgh. Requir-
ing trials that prove superiority, he says, can discourage compa-
nies from even attempting to develop new alternatives. This is al-
ready happening. Drug developers are increasingly focused on
niches where there are no good treatments to compete with, such
as rare diseases and advanced cancers. The sky’s the limit on pric-
es for these first-to-market drugs, which are often rushed through
fda approval with limited data on efficacy. Many new cancer
drugs are approved when it is shown they can shrink tumors by
30 percent, even if there is no proof that they boost survival.
This lack of meaningful data to guide patients is a major point
of Wieseler’s paper. Tunis shares her concern: with accelerated
approval, “there are more products approved, with a greater
amount of uncertainty about risks and benefits.” But there are
other solutions besides head-to-head drug trials. One idea is for
regulators and payers to require postmarket studies to track the
effectiveness of newly approved drugs—a step too often neglected.
Tunis’s center is taking another approach. Last year it helped to
convene the makers of seven experimental gene therapies for he-
mophilia with patient groups, regulators, HTA agencies and oth-
ers to agree on a set of meaningful end points for the companies’
final studies before they seek approval. Patients, for example,
asked that improvements in chronic pain and mental health be
measured along with the frequency of bleeding episodes. The
center is now looking at sickle cell therapies. If developers all use
the same outcome metrics, it will be possible to compare the var-
ious products. Patients and their doctors won’t be left in the dark.© 2019 Scientific American