SCIENCE sciencemag.org 16 AUGUST 2019 • VOL 365 ISSUE 6454 629effective (Science, 1 January 2016, p. 12).
Some had high hopes for an antibody cock-
tail named ZMapp, but a trial in 71 patients
didn’t show a significant effect on mortality.
Still, the ZMapp data were considered en-
couraging enough for researchers to use it
as a control in the new trial, named PALM,
short for the Swahili expression pamoja tu-
linde maisha, or “together save lives.”
PALM randomized patients at four Ebola
treatment centers to ZMapp, REGN-EB3,
mAb114, or remdesivir, an antiviral drug
developed by Gilead Sciences in Foster City,
California. The goal was to enroll 725 pa-
tients. But on 9 August, an independent
data and safety monitoring board reviewed
data for 499 patients and found that only
29% of patients on REGN-EB3 had died,
compared with 49% on ZMapp and 53% on
remdesivir. That result was good enough,
according to predetermined criteria, that
the trial had to end, because it’s unethical to
withhold proven treatments from patients.
The mortality rate in patients given mAb
was 34%, close enough to REGN-EB3 to
continue its use as well.
The two preparations did even better in
the 41% of trial participants who had lower
levels of Ebola virus in their blood because
they had sought treatment early. In those
people, mortality plummeted to 6% in the
REGN-EB3 group and to 11% with mAb114.
(The numbers for ZMapp and remdesivir
were 24% and 33%, respectively.) Patients
with a high viral load faced a bleaker out-
look, however: a death rate of 60% or more.
Muyembe hopes these numbers will per-
suade future patients to seek help early.
The success is remarkable given that the
trial took place in a remote, conflict-torn
area—“the worst possible conditions,” says
Anthony Fauci, who heads NIAID, a partner
in the trial. “And the beauty of it is that you
can now immediately apply it in the field.”
Indeed, patients in the four treatment
centers will now be randomized either to
REGN-EB3 or mAb114, in an extension of the
original trial. Under an ethical framework
developed by WHO, patients in all other
DRC treatment centers will also be eligible
to receive one of the two new treatments,
even though they are not yet licensed. Re-
generon and Ridgeback Biotherapeutics in
Miami, Florida, which produces mAb114,
have stockpiled enough doses to cover all
patients, Fauci says.
Muyembe, who helped discover Ebola
during the first outbreak in 1976 and tried to
save Kikwit patients using survivors’ blood
in 1995, said he never thought he’d live to
see the announcement of such hopeful data.
“Today, we have started a new chapter,” he
told reporters. “From now on, we will no
longer say that Ebola is not curable.” j
SCIENTIFIC PUBLISHINGT
he long-standing debate over open
access to research results has been
marked by a geographic divide. In
Europe, some public funders have
launched a high-profile open-access
initiative, dubbed Plan S, that would
ultimately require grantees to publish only
in journals that immediately make papers
free to all. But in the United States, federal
agencies have stuck to a decade-old policy
that allows grantees to publish in journals
that keep papers behind a paywall for up
to 1 year. Now, the divide is starting to blur,
with one prominent U.S. research program
starting to require immediate
open access to the peer-re-
viewed publications it funds.
The policy is part of the
Cancer Moonshot program at
the National Cancer Institute
(NCI) in Bethesda, Maryland,
the 7-year, $1.8 billion re-
search initiative spearheaded
in 2016 by then–Vice Presi-
dent Joe Biden after his son
Beau died of brain cancer.(The agency later realized this paper hadn’t
been funded by the moonshot.) At first,
NCI considered accepting two free shar-
ing strategies allowed by many paywalled
journals: Researchers can either publish a
draft manuscript in an online preprint re-
pository or post the final accepted paper
on their own website. But officials decided
that “people aren’t going to look at every-
body else’s website” for papers, Singer says.
Instead, NCI opted for a “strictly” open-
access policy, with rare exceptions.
The policy says moonshot authors can
publish in either a fully open-access jour-
nal or a hybrid journal that publishes both
free and paywalled papers. The authors
can also include in their
grant budgets the fees that
open-access journals charge
for peer review and other
costs, usually about $
to $3000 per paper.
NCI says it wants to en-
able grantees to publish in
highly selective journals,
including Nature, Science,
and Cell, that don’t offer
open access, and it has been
discussing options with such journals. Cell
says it expects to negotiate ways to make
moonshot papers free, and Nature says it
will consider the issue on a case-by-case
basis. Science says it will allow authors to
post accepted moonshot manuscripts on
PubMedCentral as soon as the papers are
published, and will make the papers im-
mediately open access, as part of its policy
of making papers “related to public health”
freely available.
Singer says that, for now, NCI won’t
expand the moonshot’s open-access re-
quirement to other programs run by the
$5.7 billion institute. “We consider this a
pilot program and depending on [its] suc-
cess ... we’ll determine the next steps,” she
says. But Heather Joseph, executive director
of the Scholarly Publishing and Academic
Resources Coalition in Washington, D.C.,
hopes the agency will go further. “It’s great
that this policy is in effect for this small
slice of public research,” she says. “But obvi-
ously we’d like to see it for all research.” jBy Jocelyn Kaiser“People aren’t
going to look at
everybody else’s
website” for
moonshot papers.
Dinah Singer,
National Cancer InstituteOpen access takes root at
National Cancer Institute
Cancer Moonshot program requires grantees to provide
immediate free access to papers they publish
Biden felt that broader data sharing would
speed cancer research, and after hearing
from open-access advocates he backed the
concept for all cancer research papers. NCI
officials then drafted rules that require
moonshot grantees to submit a plan for
making their publications “immediately
and broadly available to the public.”
That is a big change from the current
policy at the National Institutes of Health,
NCI’s parent agency. NIH requires only
that final papers be available through
NIH’s full-text PubMedCentral site within
12 months of publication—a delay that
publishers cherish, saying that it safe-
guards subscription revenues and keeps
journals viable.
NCI began to flesh out the moonshot
open-access policy this summer after learn-
ing that what it believed was a moonshot
paper—one of perhaps two dozen so far—
had appeared in Cell, which is paywalled,
says NCI Deputy Director for Scientific
Strategy and Development Dinah Singer.