[Ip-health] WSJ: Drugmakers Turn to Data Mining to Avoid Expensive, Lengthy Drug Trials

Thiru Balasubramaniam thiru at keionline.org
Thu Dec 26 01:20:32 PST 2019



Drugmakers Turn to Data Mining to Avoid Expensive, Lengthy Drug Trials

Pfizer, Johnson & Johnson and Amgen try to win drug approvals by analyzing
vast data sets of electronic medical records

Companies are mining hospital and doctor files for cases where patients
already took a drug in routine medical care.
By Peter Loftus
Updated Dec. 23, 2019 9:55 am ET

Drugmakers are trying to win drug approvals by parsing vast data sets of
electronic medical records, shifting away from lengthy, and costly,
clinical trials in patients.

Pfizer Inc, Johnson & Johnson and Amgen Inc. are among the drugmakers that
have submitted the data-mining analyses to the U.S. Food and Drug
Administration in seeking approval to sell new medicines or for new uses
for older ones. The FDA has approved new uses for breast cancer, bladder
cancer and leukemia drugs in part based on the data.

For the companies, the use of real-world data can cut costs and shorten
drug-development times. Instead of finding trial subjects, companies simply
mine hospital and doctor files for cases where patients already took a drug
in routine medical care, looking for changes in blood pressure, tumor size
and other readings to see if the medicine is helping or causing a side

This real-world evidence is sometimes used in lieu of a clinical trial’s
control arm, to compare outcomes for past patients who got a standard
treatment against people who are taking a new drug in a clinical trial.
Such analyses can take months, compared with years for prospective clinical

Some doctors worry about forsaking clinical trials, which are carefully
designed and conducted in patients to get a sound read on a drug’s safety
and efficacy. Health records often have errors, the skeptics say, and even
an analysis of records that are error-free doesn’t have the same scientific
rigor as a clinical trial in sizing up a drug.

Data from a clinical trial “are invaluable because it gives you a sense of,
‘In perfect circumstances, did the intervention work?’” said Dr. Joseph
Ross, professor of medicine and public health at Yale School of Medicine.

Supporters of real-world analyses say it is unlikely they would replace
standard clinical trials needed for the initial approval of a new drug.
Instead, data-mining could augment such trials or be done instead of trials
for secondary approvals of a new use, when patients are already taking the
drug for an unapproved use and can be tracked.

Clinical trials have been a bedrock of medical testing for decades. The
most reliable ones randomly assign some subjects to take an experimental
drug, while other subjects get a placebo or older drug.

But for rare diseases especially, it can take a while to even enroll enough
patients in studies. And their cost can limit the number of trials that
companies can fund, drugmakers say.

A 2016 law required the FDA to explore greater use of real-world data, and
the agency is developing standards to assess the reliability of different
data sources and which kinds of decisions the data support.

“Real-world evidence should not be a means toward dropping standards, but
rather a mechanism to have more efficiency in evidence generation while
maintaining standards,” said FDA Principal Deputy Commissioner Amy
Abernethy, a former executive at health-data firm Flatiron Health.

A market has emerged in recent years for digital drug-use information.
Iqvia Inc., which tracks prescription and health data, has about a dozen
projects under way, said Nancy Dreyer, the company’s chief scientific
officer of real-world evidence.

Tech giants like Amazon and Apple are expanding their businesses to include
electronic health records—which contain data on diagnoses, prescriptions
and other medical information. That’s creating both opportunities and
spurring privacy concerns. Here’s what to know. Photo Composite: Heather
Seidel/ The Wall Street Journal

Swiss drugmaker Roche Holding AG RHHBY -0.04% has spent about $5 billion in
recent years acquiring two health-data companies: Flatiron provides
electronic health-record software to cancer clinics and analyzes data from
those records for drug companies, while Foundation Medicine tests patients’
tumor samples for genetic mutations and aggregates the information for drug

Roche included a Flatiron analysis of patients whose cancer has a certain
genetic defect in the company’s application to market a new drug targeting
that defect, Rozlytrek, which the FDA approved in August.

Amgen used real-world data from leukemia patients to serve as a comparison
for a small clinical trial in which all patients in the trial received its
leukemia drug Blincyto. The FDA last year used the analysis in its decision
to approve a new use for Blincyto treating patients who are in remission
but have some cancerous cells that put them at risk for relapse.

Amgen would have had to enroll at least 50% more patients in the clinical
trial to have a standard control arm, said Elliott Levy, Amgen’s senior
vice president of global development.

Such trials “typically involve the potential to be randomized to existing
standard of care or even no therapy, when what patients want is the
opportunity to be treated with a promising experimental” drug, he said.

J&J used Flatiron and Foundation data to find that bladder-cancer patients
with a certain genetic trait didn’t benefit significantly from certain
immune-boosting drugs. This analysis augmented a separate clinical trial in
which patients with the same genetic trait did benefit from J&J’s
experimental drug Balversa, said Craig Tendler, vice president of oncology
clinical development and global medical affairs at J&J.

J&J included the Flatiron/Foundation results in its application for
Balversa’s approval, which the FDA granted in April 2019.

Pfizer’s breast-cancer drug Ibrance was originally approved in 2015 for
women based on a clinical trial. Some doctors began prescribing the drug to
men with breast cancer, though it wasn’t approved for use in men. Pfizer
contracted with Flatiron to mine electronic health records to see how men
who got the drug fared compared with those who hadn’t.

Separately, Iqvia reviewed prescription claims and found male breast-cancer
patients stayed on Ibrance longer than those taking other drugs.

Chris Boshoff, Pfizer’s chief development officer for oncology, said the
analyses took less than a year and cost a fraction of a clinical trial,
which he estimated would have taken four or five years and cost tens of
millions of dollars.

But an FDA reviewer of Pfizer’s application said the Flatiron data provided
limited evidence of Ibrance’s effectiveness in men because of small sample
sizes, according to an FDA document obtained by The Wall Street Journal via
a public-records request. And the two patient groups weren’t well balanced
on factors like age, the FDA reviewer said.

FDA reviewers also said it was difficult to conclude from the Iqvia
analysis whether an increase in prescription duration translated into
improved survival or a delay in disease progression.

Yet the FDA approved the use of Ibrance in men, noting that the prior
clinical trials in women gave the agency the confidence the drug could help
men, supported by the real-world evidence and Pfizer’s safety-tracking

Write to Peter Loftus at peter.loftus at wsj.com

Thiru Balasubramaniam
Geneva Representative
Knowledge Ecology International
41 22 791 6727
thiru at keionline.org

More information about the Ip-health mailing list